{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.001,
      "grad_norm": 1.7063226699829102,
      "learning_rate": 4.995e-05,
      "loss": 9.6305,
      "step": 10
    },
    {
      "epoch": 0.002,
      "grad_norm": 1.467505693435669,
      "learning_rate": 4.99e-05,
      "loss": 8.8474,
      "step": 20
    },
    {
      "epoch": 0.003,
      "grad_norm": 1.3338744640350342,
      "learning_rate": 4.9850000000000006e-05,
      "loss": 8.4272,
      "step": 30
    },
    {
      "epoch": 0.004,
      "grad_norm": 1.194218635559082,
      "learning_rate": 4.9800000000000004e-05,
      "loss": 7.9969,
      "step": 40
    },
    {
      "epoch": 0.005,
      "grad_norm": 0.9542586207389832,
      "learning_rate": 4.975e-05,
      "loss": 7.8018,
      "step": 50
    },
    {
      "epoch": 0.006,
      "grad_norm": 0.8312947154045105,
      "learning_rate": 4.97e-05,
      "loss": 7.5303,
      "step": 60
    },
    {
      "epoch": 0.007,
      "grad_norm": 0.6978892683982849,
      "learning_rate": 4.965e-05,
      "loss": 7.3733,
      "step": 70
    },
    {
      "epoch": 0.008,
      "grad_norm": 0.6895764470100403,
      "learning_rate": 4.96e-05,
      "loss": 7.2434,
      "step": 80
    },
    {
      "epoch": 0.009,
      "grad_norm": 0.5555976033210754,
      "learning_rate": 4.9550000000000005e-05,
      "loss": 7.0877,
      "step": 90
    },
    {
      "epoch": 0.01,
      "grad_norm": 0.836391806602478,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 7.0338,
      "step": 100
    },
    {
      "epoch": 0.011,
      "grad_norm": 0.782464861869812,
      "learning_rate": 4.945e-05,
      "loss": 6.878,
      "step": 110
    },
    {
      "epoch": 0.012,
      "grad_norm": 1.3705933094024658,
      "learning_rate": 4.94e-05,
      "loss": 6.5874,
      "step": 120
    },
    {
      "epoch": 0.013,
      "grad_norm": 0.7560775876045227,
      "learning_rate": 4.935e-05,
      "loss": 6.4978,
      "step": 130
    },
    {
      "epoch": 0.014,
      "grad_norm": 1.3238508701324463,
      "learning_rate": 4.93e-05,
      "loss": 6.3998,
      "step": 140
    },
    {
      "epoch": 0.015,
      "grad_norm": 0.7834548950195312,
      "learning_rate": 4.9250000000000004e-05,
      "loss": 6.2838,
      "step": 150
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.762347400188446,
      "learning_rate": 4.92e-05,
      "loss": 6.0387,
      "step": 160
    },
    {
      "epoch": 0.017,
      "grad_norm": 0.7799501419067383,
      "learning_rate": 4.915e-05,
      "loss": 6.0241,
      "step": 170
    },
    {
      "epoch": 0.018,
      "grad_norm": 0.7948866486549377,
      "learning_rate": 4.91e-05,
      "loss": 5.8776,
      "step": 180
    },
    {
      "epoch": 0.019,
      "grad_norm": 0.9890483021736145,
      "learning_rate": 4.905e-05,
      "loss": 5.747,
      "step": 190
    },
    {
      "epoch": 0.02,
      "grad_norm": 0.9131263494491577,
      "learning_rate": 4.9e-05,
      "loss": 5.644,
      "step": 200
    },
    {
      "epoch": 0.021,
      "grad_norm": 1.7073436975479126,
      "learning_rate": 4.8950000000000004e-05,
      "loss": 5.778,
      "step": 210
    },
    {
      "epoch": 0.022,
      "grad_norm": 0.8059922456741333,
      "learning_rate": 4.89e-05,
      "loss": 5.4755,
      "step": 220
    },
    {
      "epoch": 0.023,
      "grad_norm": 1.2500686645507812,
      "learning_rate": 4.885e-05,
      "loss": 5.3769,
      "step": 230
    },
    {
      "epoch": 0.024,
      "grad_norm": 1.3848680257797241,
      "learning_rate": 4.88e-05,
      "loss": 5.2105,
      "step": 240
    },
    {
      "epoch": 0.025,
      "grad_norm": 1.2381746768951416,
      "learning_rate": 4.875e-05,
      "loss": 5.1444,
      "step": 250
    },
    {
      "epoch": 0.026,
      "grad_norm": 2.7005224227905273,
      "learning_rate": 4.87e-05,
      "loss": 5.1608,
      "step": 260
    },
    {
      "epoch": 0.027,
      "grad_norm": 1.1472671031951904,
      "learning_rate": 4.8650000000000003e-05,
      "loss": 4.9456,
      "step": 270
    },
    {
      "epoch": 0.028,
      "grad_norm": 1.9849270582199097,
      "learning_rate": 4.86e-05,
      "loss": 4.8466,
      "step": 280
    },
    {
      "epoch": 0.029,
      "grad_norm": 1.857001781463623,
      "learning_rate": 4.855e-05,
      "loss": 4.7323,
      "step": 290
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.6731220483779907,
      "learning_rate": 4.85e-05,
      "loss": 4.5786,
      "step": 300
    },
    {
      "epoch": 0.031,
      "grad_norm": 1.7968906164169312,
      "learning_rate": 4.845e-05,
      "loss": 4.4588,
      "step": 310
    },
    {
      "epoch": 0.032,
      "grad_norm": 1.7908226251602173,
      "learning_rate": 4.8400000000000004e-05,
      "loss": 4.3645,
      "step": 320
    },
    {
      "epoch": 0.033,
      "grad_norm": 2.538881540298462,
      "learning_rate": 4.835e-05,
      "loss": 4.1489,
      "step": 330
    },
    {
      "epoch": 0.034,
      "grad_norm": 2.306257486343384,
      "learning_rate": 4.83e-05,
      "loss": 3.9798,
      "step": 340
    },
    {
      "epoch": 0.035,
      "grad_norm": 2.1730940341949463,
      "learning_rate": 4.825e-05,
      "loss": 4.0231,
      "step": 350
    },
    {
      "epoch": 0.036,
      "grad_norm": 2.4211463928222656,
      "learning_rate": 4.82e-05,
      "loss": 3.8495,
      "step": 360
    },
    {
      "epoch": 0.037,
      "grad_norm": 2.3698794841766357,
      "learning_rate": 4.815e-05,
      "loss": 3.6977,
      "step": 370
    },
    {
      "epoch": 0.038,
      "grad_norm": 2.147799491882324,
      "learning_rate": 4.8100000000000004e-05,
      "loss": 3.8008,
      "step": 380
    },
    {
      "epoch": 0.039,
      "grad_norm": 2.3577606678009033,
      "learning_rate": 4.805e-05,
      "loss": 3.6983,
      "step": 390
    },
    {
      "epoch": 0.04,
      "grad_norm": 2.065912961959839,
      "learning_rate": 4.8e-05,
      "loss": 3.5738,
      "step": 400
    },
    {
      "epoch": 0.041,
      "grad_norm": 2.930288314819336,
      "learning_rate": 4.795e-05,
      "loss": 3.5117,
      "step": 410
    },
    {
      "epoch": 0.042,
      "grad_norm": 2.3703155517578125,
      "learning_rate": 4.79e-05,
      "loss": 3.2483,
      "step": 420
    },
    {
      "epoch": 0.043,
      "grad_norm": 2.6050736904144287,
      "learning_rate": 4.785e-05,
      "loss": 3.2342,
      "step": 430
    },
    {
      "epoch": 0.044,
      "grad_norm": 2.0790674686431885,
      "learning_rate": 4.78e-05,
      "loss": 3.1452,
      "step": 440
    },
    {
      "epoch": 0.045,
      "grad_norm": 2.2497427463531494,
      "learning_rate": 4.775e-05,
      "loss": 3.0316,
      "step": 450
    },
    {
      "epoch": 0.046,
      "grad_norm": 2.507902145385742,
      "learning_rate": 4.77e-05,
      "loss": 2.8938,
      "step": 460
    },
    {
      "epoch": 0.047,
      "grad_norm": 2.517744541168213,
      "learning_rate": 4.765e-05,
      "loss": 2.8137,
      "step": 470
    },
    {
      "epoch": 0.048,
      "grad_norm": 3.9981460571289062,
      "learning_rate": 4.76e-05,
      "loss": 2.9864,
      "step": 480
    },
    {
      "epoch": 0.049,
      "grad_norm": 2.265026569366455,
      "learning_rate": 4.755e-05,
      "loss": 2.7839,
      "step": 490
    },
    {
      "epoch": 0.05,
      "grad_norm": 2.257293701171875,
      "learning_rate": 4.75e-05,
      "loss": 2.6834,
      "step": 500
    },
    {
      "epoch": 0.051,
      "grad_norm": 2.6932270526885986,
      "learning_rate": 4.745e-05,
      "loss": 2.5755,
      "step": 510
    },
    {
      "epoch": 0.052,
      "grad_norm": 1.7177081108093262,
      "learning_rate": 4.74e-05,
      "loss": 2.425,
      "step": 520
    },
    {
      "epoch": 0.053,
      "grad_norm": 2.2452073097229004,
      "learning_rate": 4.735e-05,
      "loss": 2.5261,
      "step": 530
    },
    {
      "epoch": 0.054,
      "grad_norm": 2.2109947204589844,
      "learning_rate": 4.73e-05,
      "loss": 2.3825,
      "step": 540
    },
    {
      "epoch": 0.055,
      "grad_norm": 2.574531078338623,
      "learning_rate": 4.7249999999999997e-05,
      "loss": 2.3087,
      "step": 550
    },
    {
      "epoch": 0.056,
      "grad_norm": 2.3631017208099365,
      "learning_rate": 4.72e-05,
      "loss": 2.3099,
      "step": 560
    },
    {
      "epoch": 0.057,
      "grad_norm": 2.3809709548950195,
      "learning_rate": 4.715e-05,
      "loss": 2.3001,
      "step": 570
    },
    {
      "epoch": 0.058,
      "grad_norm": 2.0683534145355225,
      "learning_rate": 4.71e-05,
      "loss": 2.0813,
      "step": 580
    },
    {
      "epoch": 0.059,
      "grad_norm": 2.5471837520599365,
      "learning_rate": 4.705e-05,
      "loss": 2.0378,
      "step": 590
    },
    {
      "epoch": 0.06,
      "grad_norm": 2.585564374923706,
      "learning_rate": 4.7e-05,
      "loss": 2.2062,
      "step": 600
    },
    {
      "epoch": 0.061,
      "grad_norm": 2.062100648880005,
      "learning_rate": 4.695e-05,
      "loss": 1.9914,
      "step": 610
    },
    {
      "epoch": 0.062,
      "grad_norm": 2.1019210815429688,
      "learning_rate": 4.69e-05,
      "loss": 1.9635,
      "step": 620
    },
    {
      "epoch": 0.063,
      "grad_norm": 2.630436658859253,
      "learning_rate": 4.685000000000001e-05,
      "loss": 1.9123,
      "step": 630
    },
    {
      "epoch": 0.064,
      "grad_norm": 2.1028494834899902,
      "learning_rate": 4.6800000000000006e-05,
      "loss": 1.7583,
      "step": 640
    },
    {
      "epoch": 0.065,
      "grad_norm": 2.392193078994751,
      "learning_rate": 4.6750000000000005e-05,
      "loss": 1.7532,
      "step": 650
    },
    {
      "epoch": 0.066,
      "grad_norm": 2.004413366317749,
      "learning_rate": 4.6700000000000003e-05,
      "loss": 1.6978,
      "step": 660
    },
    {
      "epoch": 0.067,
      "grad_norm": 2.210513114929199,
      "learning_rate": 4.665e-05,
      "loss": 1.6311,
      "step": 670
    },
    {
      "epoch": 0.068,
      "grad_norm": 1.8464936017990112,
      "learning_rate": 4.660000000000001e-05,
      "loss": 1.5507,
      "step": 680
    },
    {
      "epoch": 0.069,
      "grad_norm": 2.0246541500091553,
      "learning_rate": 4.655000000000001e-05,
      "loss": 1.5637,
      "step": 690
    },
    {
      "epoch": 0.07,
      "grad_norm": 2.199751138687134,
      "learning_rate": 4.6500000000000005e-05,
      "loss": 1.5603,
      "step": 700
    },
    {
      "epoch": 0.071,
      "grad_norm": 2.2002196311950684,
      "learning_rate": 4.6450000000000004e-05,
      "loss": 1.4558,
      "step": 710
    },
    {
      "epoch": 0.072,
      "grad_norm": 1.7826759815216064,
      "learning_rate": 4.64e-05,
      "loss": 1.4309,
      "step": 720
    },
    {
      "epoch": 0.073,
      "grad_norm": 1.760297417640686,
      "learning_rate": 4.635e-05,
      "loss": 1.3531,
      "step": 730
    },
    {
      "epoch": 0.074,
      "grad_norm": 2.0505475997924805,
      "learning_rate": 4.630000000000001e-05,
      "loss": 1.3641,
      "step": 740
    },
    {
      "epoch": 0.075,
      "grad_norm": 2.1375396251678467,
      "learning_rate": 4.6250000000000006e-05,
      "loss": 1.3259,
      "step": 750
    },
    {
      "epoch": 0.076,
      "grad_norm": 1.8252328634262085,
      "learning_rate": 4.6200000000000005e-05,
      "loss": 1.2026,
      "step": 760
    },
    {
      "epoch": 0.077,
      "grad_norm": 1.8945906162261963,
      "learning_rate": 4.6150000000000004e-05,
      "loss": 1.2878,
      "step": 770
    },
    {
      "epoch": 0.078,
      "grad_norm": 1.7990881204605103,
      "learning_rate": 4.61e-05,
      "loss": 1.1853,
      "step": 780
    },
    {
      "epoch": 0.079,
      "grad_norm": 1.4897470474243164,
      "learning_rate": 4.605e-05,
      "loss": 1.1279,
      "step": 790
    },
    {
      "epoch": 0.08,
      "grad_norm": 2.2804617881774902,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.0804,
      "step": 800
    },
    {
      "epoch": 0.081,
      "grad_norm": 1.4800664186477661,
      "learning_rate": 4.5950000000000006e-05,
      "loss": 1.0361,
      "step": 810
    },
    {
      "epoch": 0.082,
      "grad_norm": 1.3526049852371216,
      "learning_rate": 4.5900000000000004e-05,
      "loss": 1.0585,
      "step": 820
    },
    {
      "epoch": 0.083,
      "grad_norm": 1.534173607826233,
      "learning_rate": 4.585e-05,
      "loss": 1.0206,
      "step": 830
    },
    {
      "epoch": 0.084,
      "grad_norm": 1.4844435453414917,
      "learning_rate": 4.58e-05,
      "loss": 0.9758,
      "step": 840
    },
    {
      "epoch": 0.085,
      "grad_norm": 1.533679485321045,
      "learning_rate": 4.575e-05,
      "loss": 0.9168,
      "step": 850
    },
    {
      "epoch": 0.086,
      "grad_norm": 1.456162691116333,
      "learning_rate": 4.5700000000000006e-05,
      "loss": 0.8913,
      "step": 860
    },
    {
      "epoch": 0.087,
      "grad_norm": 1.7335631847381592,
      "learning_rate": 4.5650000000000005e-05,
      "loss": 0.9154,
      "step": 870
    },
    {
      "epoch": 0.088,
      "grad_norm": 1.3331761360168457,
      "learning_rate": 4.5600000000000004e-05,
      "loss": 0.8483,
      "step": 880
    },
    {
      "epoch": 0.089,
      "grad_norm": 1.6703053712844849,
      "learning_rate": 4.555e-05,
      "loss": 0.8116,
      "step": 890
    },
    {
      "epoch": 0.09,
      "grad_norm": 1.275975227355957,
      "learning_rate": 4.55e-05,
      "loss": 0.7869,
      "step": 900
    },
    {
      "epoch": 0.091,
      "grad_norm": 1.3800309896469116,
      "learning_rate": 4.545000000000001e-05,
      "loss": 0.7637,
      "step": 910
    },
    {
      "epoch": 0.092,
      "grad_norm": 1.9472386837005615,
      "learning_rate": 4.5400000000000006e-05,
      "loss": 0.7212,
      "step": 920
    },
    {
      "epoch": 0.093,
      "grad_norm": 1.3451333045959473,
      "learning_rate": 4.5350000000000005e-05,
      "loss": 0.6829,
      "step": 930
    },
    {
      "epoch": 0.094,
      "grad_norm": 1.5209784507751465,
      "learning_rate": 4.53e-05,
      "loss": 0.729,
      "step": 940
    },
    {
      "epoch": 0.095,
      "grad_norm": 1.3944469690322876,
      "learning_rate": 4.525e-05,
      "loss": 0.6732,
      "step": 950
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.2177132368087769,
      "learning_rate": 4.52e-05,
      "loss": 0.6188,
      "step": 960
    },
    {
      "epoch": 0.097,
      "grad_norm": 1.5988528728485107,
      "learning_rate": 4.5150000000000006e-05,
      "loss": 0.6622,
      "step": 970
    },
    {
      "epoch": 0.098,
      "grad_norm": 1.3636531829833984,
      "learning_rate": 4.5100000000000005e-05,
      "loss": 0.5792,
      "step": 980
    },
    {
      "epoch": 0.099,
      "grad_norm": 1.377453088760376,
      "learning_rate": 4.5050000000000004e-05,
      "loss": 0.6062,
      "step": 990
    },
    {
      "epoch": 0.1,
      "grad_norm": 2.295713186264038,
      "learning_rate": 4.5e-05,
      "loss": 0.5709,
      "step": 1000
    },
    {
      "epoch": 0.101,
      "grad_norm": 1.35196852684021,
      "learning_rate": 4.495e-05,
      "loss": 0.5521,
      "step": 1010
    },
    {
      "epoch": 0.102,
      "grad_norm": 1.0617187023162842,
      "learning_rate": 4.49e-05,
      "loss": 0.5147,
      "step": 1020
    },
    {
      "epoch": 0.103,
      "grad_norm": 1.3035167455673218,
      "learning_rate": 4.4850000000000006e-05,
      "loss": 0.5081,
      "step": 1030
    },
    {
      "epoch": 0.104,
      "grad_norm": 1.2835568189620972,
      "learning_rate": 4.4800000000000005e-05,
      "loss": 0.5,
      "step": 1040
    },
    {
      "epoch": 0.105,
      "grad_norm": 1.0403038263320923,
      "learning_rate": 4.4750000000000004e-05,
      "loss": 0.4825,
      "step": 1050
    },
    {
      "epoch": 0.106,
      "grad_norm": 0.9538235068321228,
      "learning_rate": 4.47e-05,
      "loss": 0.4316,
      "step": 1060
    },
    {
      "epoch": 0.107,
      "grad_norm": 1.4246289730072021,
      "learning_rate": 4.465e-05,
      "loss": 0.4304,
      "step": 1070
    },
    {
      "epoch": 0.108,
      "grad_norm": 1.1217833757400513,
      "learning_rate": 4.46e-05,
      "loss": 0.4397,
      "step": 1080
    },
    {
      "epoch": 0.109,
      "grad_norm": 1.0411335229873657,
      "learning_rate": 4.4550000000000005e-05,
      "loss": 0.4057,
      "step": 1090
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.8498069643974304,
      "learning_rate": 4.4500000000000004e-05,
      "loss": 0.3933,
      "step": 1100
    },
    {
      "epoch": 0.111,
      "grad_norm": 1.1270406246185303,
      "learning_rate": 4.445e-05,
      "loss": 0.366,
      "step": 1110
    },
    {
      "epoch": 0.112,
      "grad_norm": 1.189041256904602,
      "learning_rate": 4.44e-05,
      "loss": 0.3407,
      "step": 1120
    },
    {
      "epoch": 0.113,
      "grad_norm": 0.9837467670440674,
      "learning_rate": 4.435e-05,
      "loss": 0.3511,
      "step": 1130
    },
    {
      "epoch": 0.114,
      "grad_norm": 1.0432955026626587,
      "learning_rate": 4.43e-05,
      "loss": 0.3381,
      "step": 1140
    },
    {
      "epoch": 0.115,
      "grad_norm": 0.9529951810836792,
      "learning_rate": 4.4250000000000005e-05,
      "loss": 0.3189,
      "step": 1150
    },
    {
      "epoch": 0.116,
      "grad_norm": 1.008836030960083,
      "learning_rate": 4.4200000000000004e-05,
      "loss": 0.3077,
      "step": 1160
    },
    {
      "epoch": 0.117,
      "grad_norm": 1.0005086660385132,
      "learning_rate": 4.415e-05,
      "loss": 0.3001,
      "step": 1170
    },
    {
      "epoch": 0.118,
      "grad_norm": 1.1065175533294678,
      "learning_rate": 4.41e-05,
      "loss": 0.28,
      "step": 1180
    },
    {
      "epoch": 0.119,
      "grad_norm": 0.6701949834823608,
      "learning_rate": 4.405e-05,
      "loss": 0.2692,
      "step": 1190
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.7154658436775208,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2663,
      "step": 1200
    },
    {
      "epoch": 0.121,
      "grad_norm": 0.6997113823890686,
      "learning_rate": 4.3950000000000004e-05,
      "loss": 0.2595,
      "step": 1210
    },
    {
      "epoch": 0.122,
      "grad_norm": 0.9047608971595764,
      "learning_rate": 4.39e-05,
      "loss": 0.2558,
      "step": 1220
    },
    {
      "epoch": 0.123,
      "grad_norm": 0.8508415222167969,
      "learning_rate": 4.385e-05,
      "loss": 0.2459,
      "step": 1230
    },
    {
      "epoch": 0.124,
      "grad_norm": 0.6505220532417297,
      "learning_rate": 4.38e-05,
      "loss": 0.2236,
      "step": 1240
    },
    {
      "epoch": 0.125,
      "grad_norm": 0.5360460877418518,
      "learning_rate": 4.375e-05,
      "loss": 0.2189,
      "step": 1250
    },
    {
      "epoch": 0.126,
      "grad_norm": 0.560817539691925,
      "learning_rate": 4.3700000000000005e-05,
      "loss": 0.2166,
      "step": 1260
    },
    {
      "epoch": 0.127,
      "grad_norm": 0.7089666128158569,
      "learning_rate": 4.3650000000000004e-05,
      "loss": 0.2026,
      "step": 1270
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.5265817046165466,
      "learning_rate": 4.36e-05,
      "loss": 0.197,
      "step": 1280
    },
    {
      "epoch": 0.129,
      "grad_norm": 0.6629377007484436,
      "learning_rate": 4.355e-05,
      "loss": 0.1934,
      "step": 1290
    },
    {
      "epoch": 0.13,
      "grad_norm": 1.0730735063552856,
      "learning_rate": 4.35e-05,
      "loss": 0.1807,
      "step": 1300
    },
    {
      "epoch": 0.131,
      "grad_norm": 0.6990699172019958,
      "learning_rate": 4.345e-05,
      "loss": 0.1845,
      "step": 1310
    },
    {
      "epoch": 0.132,
      "grad_norm": 0.5047340393066406,
      "learning_rate": 4.3400000000000005e-05,
      "loss": 0.1725,
      "step": 1320
    },
    {
      "epoch": 0.133,
      "grad_norm": 0.6830994486808777,
      "learning_rate": 4.335e-05,
      "loss": 0.1687,
      "step": 1330
    },
    {
      "epoch": 0.134,
      "grad_norm": 0.5861710906028748,
      "learning_rate": 4.33e-05,
      "loss": 0.1671,
      "step": 1340
    },
    {
      "epoch": 0.135,
      "grad_norm": 0.43594300746917725,
      "learning_rate": 4.325e-05,
      "loss": 0.1467,
      "step": 1350
    },
    {
      "epoch": 0.136,
      "grad_norm": 0.44587692618370056,
      "learning_rate": 4.32e-05,
      "loss": 0.1509,
      "step": 1360
    },
    {
      "epoch": 0.137,
      "grad_norm": 0.5523977875709534,
      "learning_rate": 4.315e-05,
      "loss": 0.1434,
      "step": 1370
    },
    {
      "epoch": 0.138,
      "grad_norm": 0.6139170527458191,
      "learning_rate": 4.3100000000000004e-05,
      "loss": 0.1433,
      "step": 1380
    },
    {
      "epoch": 0.139,
      "grad_norm": 0.6169497966766357,
      "learning_rate": 4.305e-05,
      "loss": 0.1365,
      "step": 1390
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.49120134115219116,
      "learning_rate": 4.3e-05,
      "loss": 0.1287,
      "step": 1400
    },
    {
      "epoch": 0.141,
      "grad_norm": 0.451753169298172,
      "learning_rate": 4.295e-05,
      "loss": 0.1142,
      "step": 1410
    },
    {
      "epoch": 0.142,
      "grad_norm": 0.5429627895355225,
      "learning_rate": 4.29e-05,
      "loss": 0.134,
      "step": 1420
    },
    {
      "epoch": 0.143,
      "grad_norm": 0.7613041400909424,
      "learning_rate": 4.285e-05,
      "loss": 0.1391,
      "step": 1430
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.4953358471393585,
      "learning_rate": 4.2800000000000004e-05,
      "loss": 0.1197,
      "step": 1440
    },
    {
      "epoch": 0.145,
      "grad_norm": 0.3657626509666443,
      "learning_rate": 4.275e-05,
      "loss": 0.1071,
      "step": 1450
    },
    {
      "epoch": 0.146,
      "grad_norm": 0.44240206480026245,
      "learning_rate": 4.27e-05,
      "loss": 0.1111,
      "step": 1460
    },
    {
      "epoch": 0.147,
      "grad_norm": 0.5007165670394897,
      "learning_rate": 4.265e-05,
      "loss": 0.1056,
      "step": 1470
    },
    {
      "epoch": 0.148,
      "grad_norm": 0.4580256938934326,
      "learning_rate": 4.26e-05,
      "loss": 0.1049,
      "step": 1480
    },
    {
      "epoch": 0.149,
      "grad_norm": 0.4970822036266327,
      "learning_rate": 4.2550000000000004e-05,
      "loss": 0.1032,
      "step": 1490
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4138182997703552,
      "learning_rate": 4.25e-05,
      "loss": 0.0961,
      "step": 1500
    },
    {
      "epoch": 0.151,
      "grad_norm": 0.4013712406158447,
      "learning_rate": 4.245e-05,
      "loss": 0.0949,
      "step": 1510
    },
    {
      "epoch": 0.152,
      "grad_norm": 0.3868940770626068,
      "learning_rate": 4.24e-05,
      "loss": 0.0837,
      "step": 1520
    },
    {
      "epoch": 0.153,
      "grad_norm": 0.3113015294075012,
      "learning_rate": 4.235e-05,
      "loss": 0.0909,
      "step": 1530
    },
    {
      "epoch": 0.154,
      "grad_norm": 0.3569623529911041,
      "learning_rate": 4.23e-05,
      "loss": 0.0908,
      "step": 1540
    },
    {
      "epoch": 0.155,
      "grad_norm": 0.3841746151447296,
      "learning_rate": 4.2250000000000004e-05,
      "loss": 0.0806,
      "step": 1550
    },
    {
      "epoch": 0.156,
      "grad_norm": 0.6565550565719604,
      "learning_rate": 4.22e-05,
      "loss": 0.075,
      "step": 1560
    },
    {
      "epoch": 0.157,
      "grad_norm": 0.4816874563694,
      "learning_rate": 4.215e-05,
      "loss": 0.0858,
      "step": 1570
    },
    {
      "epoch": 0.158,
      "grad_norm": 0.30408933758735657,
      "learning_rate": 4.21e-05,
      "loss": 0.0704,
      "step": 1580
    },
    {
      "epoch": 0.159,
      "grad_norm": 0.43388792872428894,
      "learning_rate": 4.205e-05,
      "loss": 0.0671,
      "step": 1590
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.33304253220558167,
      "learning_rate": 4.2e-05,
      "loss": 0.07,
      "step": 1600
    },
    {
      "epoch": 0.161,
      "grad_norm": 0.4260387420654297,
      "learning_rate": 4.195e-05,
      "loss": 0.0691,
      "step": 1610
    },
    {
      "epoch": 0.162,
      "grad_norm": 0.37930798530578613,
      "learning_rate": 4.19e-05,
      "loss": 0.0715,
      "step": 1620
    },
    {
      "epoch": 0.163,
      "grad_norm": 0.3198983669281006,
      "learning_rate": 4.185e-05,
      "loss": 0.0651,
      "step": 1630
    },
    {
      "epoch": 0.164,
      "grad_norm": 0.3510359823703766,
      "learning_rate": 4.18e-05,
      "loss": 0.058,
      "step": 1640
    },
    {
      "epoch": 0.165,
      "grad_norm": 0.41047966480255127,
      "learning_rate": 4.175e-05,
      "loss": 0.065,
      "step": 1650
    },
    {
      "epoch": 0.166,
      "grad_norm": 0.3054174482822418,
      "learning_rate": 4.17e-05,
      "loss": 0.0564,
      "step": 1660
    },
    {
      "epoch": 0.167,
      "grad_norm": 0.29319772124290466,
      "learning_rate": 4.165e-05,
      "loss": 0.0599,
      "step": 1670
    },
    {
      "epoch": 0.168,
      "grad_norm": 0.257354736328125,
      "learning_rate": 4.16e-05,
      "loss": 0.0536,
      "step": 1680
    },
    {
      "epoch": 0.169,
      "grad_norm": 0.25215694308280945,
      "learning_rate": 4.155e-05,
      "loss": 0.0587,
      "step": 1690
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4573931097984314,
      "learning_rate": 4.15e-05,
      "loss": 0.0524,
      "step": 1700
    },
    {
      "epoch": 0.171,
      "grad_norm": 0.3514876663684845,
      "learning_rate": 4.145e-05,
      "loss": 0.0551,
      "step": 1710
    },
    {
      "epoch": 0.172,
      "grad_norm": 0.3239930272102356,
      "learning_rate": 4.14e-05,
      "loss": 0.0499,
      "step": 1720
    },
    {
      "epoch": 0.173,
      "grad_norm": 0.20213039219379425,
      "learning_rate": 4.135e-05,
      "loss": 0.0521,
      "step": 1730
    },
    {
      "epoch": 0.174,
      "grad_norm": 0.21831783652305603,
      "learning_rate": 4.13e-05,
      "loss": 0.0469,
      "step": 1740
    },
    {
      "epoch": 0.175,
      "grad_norm": 0.2585163712501526,
      "learning_rate": 4.125e-05,
      "loss": 0.0469,
      "step": 1750
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.21717113256454468,
      "learning_rate": 4.12e-05,
      "loss": 0.0455,
      "step": 1760
    },
    {
      "epoch": 0.177,
      "grad_norm": 0.27248838543891907,
      "learning_rate": 4.115e-05,
      "loss": 0.046,
      "step": 1770
    },
    {
      "epoch": 0.178,
      "grad_norm": 0.2503461241722107,
      "learning_rate": 4.11e-05,
      "loss": 0.0447,
      "step": 1780
    },
    {
      "epoch": 0.179,
      "grad_norm": 0.27404382824897766,
      "learning_rate": 4.105e-05,
      "loss": 0.0437,
      "step": 1790
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.23549066483974457,
      "learning_rate": 4.1e-05,
      "loss": 0.0423,
      "step": 1800
    },
    {
      "epoch": 0.181,
      "grad_norm": 0.19369937479496002,
      "learning_rate": 4.095e-05,
      "loss": 0.0408,
      "step": 1810
    },
    {
      "epoch": 0.182,
      "grad_norm": 0.20560242235660553,
      "learning_rate": 4.09e-05,
      "loss": 0.0379,
      "step": 1820
    },
    {
      "epoch": 0.183,
      "grad_norm": 0.34989863634109497,
      "learning_rate": 4.085e-05,
      "loss": 0.0364,
      "step": 1830
    },
    {
      "epoch": 0.184,
      "grad_norm": 0.2310326248407364,
      "learning_rate": 4.08e-05,
      "loss": 0.0385,
      "step": 1840
    },
    {
      "epoch": 0.185,
      "grad_norm": 0.21055462956428528,
      "learning_rate": 4.075e-05,
      "loss": 0.0351,
      "step": 1850
    },
    {
      "epoch": 0.186,
      "grad_norm": 0.3251895308494568,
      "learning_rate": 4.07e-05,
      "loss": 0.0381,
      "step": 1860
    },
    {
      "epoch": 0.187,
      "grad_norm": 0.2887445390224457,
      "learning_rate": 4.065e-05,
      "loss": 0.0341,
      "step": 1870
    },
    {
      "epoch": 0.188,
      "grad_norm": 0.15948843955993652,
      "learning_rate": 4.0600000000000004e-05,
      "loss": 0.0313,
      "step": 1880
    },
    {
      "epoch": 0.189,
      "grad_norm": 0.2413359135389328,
      "learning_rate": 4.055e-05,
      "loss": 0.0338,
      "step": 1890
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.2132706493139267,
      "learning_rate": 4.05e-05,
      "loss": 0.0339,
      "step": 1900
    },
    {
      "epoch": 0.191,
      "grad_norm": 0.17968431115150452,
      "learning_rate": 4.045000000000001e-05,
      "loss": 0.0317,
      "step": 1910
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.15828929841518402,
      "learning_rate": 4.0400000000000006e-05,
      "loss": 0.0302,
      "step": 1920
    },
    {
      "epoch": 0.193,
      "grad_norm": 0.18106874823570251,
      "learning_rate": 4.0350000000000005e-05,
      "loss": 0.0331,
      "step": 1930
    },
    {
      "epoch": 0.194,
      "grad_norm": 0.34827324748039246,
      "learning_rate": 4.0300000000000004e-05,
      "loss": 0.032,
      "step": 1940
    },
    {
      "epoch": 0.195,
      "grad_norm": 0.21621111035346985,
      "learning_rate": 4.025e-05,
      "loss": 0.0317,
      "step": 1950
    },
    {
      "epoch": 0.196,
      "grad_norm": 0.2159423679113388,
      "learning_rate": 4.02e-05,
      "loss": 0.0296,
      "step": 1960
    },
    {
      "epoch": 0.197,
      "grad_norm": 0.17750391364097595,
      "learning_rate": 4.015000000000001e-05,
      "loss": 0.0297,
      "step": 1970
    },
    {
      "epoch": 0.198,
      "grad_norm": 0.13952311873435974,
      "learning_rate": 4.0100000000000006e-05,
      "loss": 0.0279,
      "step": 1980
    },
    {
      "epoch": 0.199,
      "grad_norm": 0.19622887670993805,
      "learning_rate": 4.0050000000000004e-05,
      "loss": 0.0278,
      "step": 1990
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.14959514141082764,
      "learning_rate": 4e-05,
      "loss": 0.0251,
      "step": 2000
    },
    {
      "epoch": 0.201,
      "grad_norm": 0.17456738650798798,
      "learning_rate": 3.995e-05,
      "loss": 0.0315,
      "step": 2010
    },
    {
      "epoch": 0.202,
      "grad_norm": 0.15893588960170746,
      "learning_rate": 3.99e-05,
      "loss": 0.0243,
      "step": 2020
    },
    {
      "epoch": 0.203,
      "grad_norm": 0.14638105034828186,
      "learning_rate": 3.9850000000000006e-05,
      "loss": 0.0247,
      "step": 2030
    },
    {
      "epoch": 0.204,
      "grad_norm": 0.1714017242193222,
      "learning_rate": 3.9800000000000005e-05,
      "loss": 0.0252,
      "step": 2040
    },
    {
      "epoch": 0.205,
      "grad_norm": 0.18679572641849518,
      "learning_rate": 3.9750000000000004e-05,
      "loss": 0.0234,
      "step": 2050
    },
    {
      "epoch": 0.206,
      "grad_norm": 0.10623681545257568,
      "learning_rate": 3.97e-05,
      "loss": 0.0256,
      "step": 2060
    },
    {
      "epoch": 0.207,
      "grad_norm": 0.18566076457500458,
      "learning_rate": 3.965e-05,
      "loss": 0.0238,
      "step": 2070
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.12487553805112839,
      "learning_rate": 3.960000000000001e-05,
      "loss": 0.0226,
      "step": 2080
    },
    {
      "epoch": 0.209,
      "grad_norm": 0.13191473484039307,
      "learning_rate": 3.9550000000000006e-05,
      "loss": 0.0232,
      "step": 2090
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.225613072514534,
      "learning_rate": 3.9500000000000005e-05,
      "loss": 0.0226,
      "step": 2100
    },
    {
      "epoch": 0.211,
      "grad_norm": 0.10896781831979752,
      "learning_rate": 3.9450000000000003e-05,
      "loss": 0.0206,
      "step": 2110
    },
    {
      "epoch": 0.212,
      "grad_norm": 0.16153796017169952,
      "learning_rate": 3.94e-05,
      "loss": 0.0195,
      "step": 2120
    },
    {
      "epoch": 0.213,
      "grad_norm": 0.19171251356601715,
      "learning_rate": 3.935e-05,
      "loss": 0.0203,
      "step": 2130
    },
    {
      "epoch": 0.214,
      "grad_norm": 0.13199982047080994,
      "learning_rate": 3.9300000000000007e-05,
      "loss": 0.0194,
      "step": 2140
    },
    {
      "epoch": 0.215,
      "grad_norm": 0.12839478254318237,
      "learning_rate": 3.9250000000000005e-05,
      "loss": 0.0227,
      "step": 2150
    },
    {
      "epoch": 0.216,
      "grad_norm": 0.10787441581487656,
      "learning_rate": 3.9200000000000004e-05,
      "loss": 0.0195,
      "step": 2160
    },
    {
      "epoch": 0.217,
      "grad_norm": 0.1551046371459961,
      "learning_rate": 3.915e-05,
      "loss": 0.019,
      "step": 2170
    },
    {
      "epoch": 0.218,
      "grad_norm": 0.18844197690486908,
      "learning_rate": 3.91e-05,
      "loss": 0.02,
      "step": 2180
    },
    {
      "epoch": 0.219,
      "grad_norm": 0.21247665584087372,
      "learning_rate": 3.905e-05,
      "loss": 0.0206,
      "step": 2190
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.11881183087825775,
      "learning_rate": 3.9000000000000006e-05,
      "loss": 0.0176,
      "step": 2200
    },
    {
      "epoch": 0.221,
      "grad_norm": 0.16291823983192444,
      "learning_rate": 3.8950000000000005e-05,
      "loss": 0.0178,
      "step": 2210
    },
    {
      "epoch": 0.222,
      "grad_norm": 0.14063787460327148,
      "learning_rate": 3.8900000000000004e-05,
      "loss": 0.0177,
      "step": 2220
    },
    {
      "epoch": 0.223,
      "grad_norm": 0.15583930909633636,
      "learning_rate": 3.885e-05,
      "loss": 0.0185,
      "step": 2230
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.16128882765769958,
      "learning_rate": 3.88e-05,
      "loss": 0.0168,
      "step": 2240
    },
    {
      "epoch": 0.225,
      "grad_norm": 0.1588473916053772,
      "learning_rate": 3.875e-05,
      "loss": 0.0167,
      "step": 2250
    },
    {
      "epoch": 0.226,
      "grad_norm": 0.10487533360719681,
      "learning_rate": 3.8700000000000006e-05,
      "loss": 0.0154,
      "step": 2260
    },
    {
      "epoch": 0.227,
      "grad_norm": 0.2638506591320038,
      "learning_rate": 3.8650000000000004e-05,
      "loss": 0.0179,
      "step": 2270
    },
    {
      "epoch": 0.228,
      "grad_norm": 0.12504911422729492,
      "learning_rate": 3.86e-05,
      "loss": 0.016,
      "step": 2280
    },
    {
      "epoch": 0.229,
      "grad_norm": 0.11655262857675552,
      "learning_rate": 3.855e-05,
      "loss": 0.0164,
      "step": 2290
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.10052930563688278,
      "learning_rate": 3.85e-05,
      "loss": 0.0143,
      "step": 2300
    },
    {
      "epoch": 0.231,
      "grad_norm": 0.07682032138109207,
      "learning_rate": 3.845e-05,
      "loss": 0.0165,
      "step": 2310
    },
    {
      "epoch": 0.232,
      "grad_norm": 0.12146533280611038,
      "learning_rate": 3.8400000000000005e-05,
      "loss": 0.0147,
      "step": 2320
    },
    {
      "epoch": 0.233,
      "grad_norm": 0.16349685192108154,
      "learning_rate": 3.8350000000000004e-05,
      "loss": 0.0166,
      "step": 2330
    },
    {
      "epoch": 0.234,
      "grad_norm": 0.10822432488203049,
      "learning_rate": 3.83e-05,
      "loss": 0.0146,
      "step": 2340
    },
    {
      "epoch": 0.235,
      "grad_norm": 0.0805143415927887,
      "learning_rate": 3.825e-05,
      "loss": 0.0131,
      "step": 2350
    },
    {
      "epoch": 0.236,
      "grad_norm": 0.08285068720579147,
      "learning_rate": 3.82e-05,
      "loss": 0.0134,
      "step": 2360
    },
    {
      "epoch": 0.237,
      "grad_norm": 0.0886882022023201,
      "learning_rate": 3.8150000000000006e-05,
      "loss": 0.0133,
      "step": 2370
    },
    {
      "epoch": 0.238,
      "grad_norm": 0.08793161064386368,
      "learning_rate": 3.8100000000000005e-05,
      "loss": 0.0142,
      "step": 2380
    },
    {
      "epoch": 0.239,
      "grad_norm": 0.08325997740030289,
      "learning_rate": 3.805e-05,
      "loss": 0.012,
      "step": 2390
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.10990972816944122,
      "learning_rate": 3.8e-05,
      "loss": 0.0134,
      "step": 2400
    },
    {
      "epoch": 0.241,
      "grad_norm": 0.06695697456598282,
      "learning_rate": 3.795e-05,
      "loss": 0.0119,
      "step": 2410
    },
    {
      "epoch": 0.242,
      "grad_norm": 0.08304648846387863,
      "learning_rate": 3.79e-05,
      "loss": 0.0128,
      "step": 2420
    },
    {
      "epoch": 0.243,
      "grad_norm": 0.09513472020626068,
      "learning_rate": 3.7850000000000005e-05,
      "loss": 0.0146,
      "step": 2430
    },
    {
      "epoch": 0.244,
      "grad_norm": 0.07892587035894394,
      "learning_rate": 3.7800000000000004e-05,
      "loss": 0.0116,
      "step": 2440
    },
    {
      "epoch": 0.245,
      "grad_norm": 0.12630197405815125,
      "learning_rate": 3.775e-05,
      "loss": 0.0132,
      "step": 2450
    },
    {
      "epoch": 0.246,
      "grad_norm": 0.08250122517347336,
      "learning_rate": 3.77e-05,
      "loss": 0.013,
      "step": 2460
    },
    {
      "epoch": 0.247,
      "grad_norm": 0.09903154522180557,
      "learning_rate": 3.765e-05,
      "loss": 0.0117,
      "step": 2470
    },
    {
      "epoch": 0.248,
      "grad_norm": 0.09059987217187881,
      "learning_rate": 3.76e-05,
      "loss": 0.0111,
      "step": 2480
    },
    {
      "epoch": 0.249,
      "grad_norm": 0.09777077287435532,
      "learning_rate": 3.7550000000000005e-05,
      "loss": 0.0142,
      "step": 2490
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.1801980435848236,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0121,
      "step": 2500
    },
    {
      "epoch": 0.251,
      "grad_norm": 0.08936703950166702,
      "learning_rate": 3.745e-05,
      "loss": 0.0112,
      "step": 2510
    },
    {
      "epoch": 0.252,
      "grad_norm": 0.09601296484470367,
      "learning_rate": 3.74e-05,
      "loss": 0.0116,
      "step": 2520
    },
    {
      "epoch": 0.253,
      "grad_norm": 0.08924739062786102,
      "learning_rate": 3.735e-05,
      "loss": 0.0119,
      "step": 2530
    },
    {
      "epoch": 0.254,
      "grad_norm": 0.07558383047580719,
      "learning_rate": 3.73e-05,
      "loss": 0.0108,
      "step": 2540
    },
    {
      "epoch": 0.255,
      "grad_norm": 0.05701779946684837,
      "learning_rate": 3.7250000000000004e-05,
      "loss": 0.011,
      "step": 2550
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.0955379456281662,
      "learning_rate": 3.72e-05,
      "loss": 0.0112,
      "step": 2560
    },
    {
      "epoch": 0.257,
      "grad_norm": 0.07837249338626862,
      "learning_rate": 3.715e-05,
      "loss": 0.0111,
      "step": 2570
    },
    {
      "epoch": 0.258,
      "grad_norm": 0.09438953548669815,
      "learning_rate": 3.71e-05,
      "loss": 0.0121,
      "step": 2580
    },
    {
      "epoch": 0.259,
      "grad_norm": 0.08802532404661179,
      "learning_rate": 3.705e-05,
      "loss": 0.0098,
      "step": 2590
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.0785641148686409,
      "learning_rate": 3.7e-05,
      "loss": 0.0106,
      "step": 2600
    },
    {
      "epoch": 0.261,
      "grad_norm": 0.10036404430866241,
      "learning_rate": 3.6950000000000004e-05,
      "loss": 0.0107,
      "step": 2610
    },
    {
      "epoch": 0.262,
      "grad_norm": 0.0663432776927948,
      "learning_rate": 3.69e-05,
      "loss": 0.0098,
      "step": 2620
    },
    {
      "epoch": 0.263,
      "grad_norm": 0.06886564195156097,
      "learning_rate": 3.685e-05,
      "loss": 0.0094,
      "step": 2630
    },
    {
      "epoch": 0.264,
      "grad_norm": 0.06641737371683121,
      "learning_rate": 3.68e-05,
      "loss": 0.0112,
      "step": 2640
    },
    {
      "epoch": 0.265,
      "grad_norm": 0.1470363438129425,
      "learning_rate": 3.675e-05,
      "loss": 0.0118,
      "step": 2650
    },
    {
      "epoch": 0.266,
      "grad_norm": 0.08694775402545929,
      "learning_rate": 3.6700000000000004e-05,
      "loss": 0.0105,
      "step": 2660
    },
    {
      "epoch": 0.267,
      "grad_norm": 0.08168693631887436,
      "learning_rate": 3.665e-05,
      "loss": 0.0101,
      "step": 2670
    },
    {
      "epoch": 0.268,
      "grad_norm": 0.06114206463098526,
      "learning_rate": 3.66e-05,
      "loss": 0.0097,
      "step": 2680
    },
    {
      "epoch": 0.269,
      "grad_norm": 0.09011895209550858,
      "learning_rate": 3.655e-05,
      "loss": 0.0101,
      "step": 2690
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.06499819457530975,
      "learning_rate": 3.65e-05,
      "loss": 0.0089,
      "step": 2700
    },
    {
      "epoch": 0.271,
      "grad_norm": 0.08157055824995041,
      "learning_rate": 3.645e-05,
      "loss": 0.0099,
      "step": 2710
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.06255003809928894,
      "learning_rate": 3.6400000000000004e-05,
      "loss": 0.0091,
      "step": 2720
    },
    {
      "epoch": 0.273,
      "grad_norm": 0.13641099631786346,
      "learning_rate": 3.635e-05,
      "loss": 0.01,
      "step": 2730
    },
    {
      "epoch": 0.274,
      "grad_norm": 0.06449054926633835,
      "learning_rate": 3.63e-05,
      "loss": 0.0087,
      "step": 2740
    },
    {
      "epoch": 0.275,
      "grad_norm": 0.09242594987154007,
      "learning_rate": 3.625e-05,
      "loss": 0.0084,
      "step": 2750
    },
    {
      "epoch": 0.276,
      "grad_norm": 0.14216932654380798,
      "learning_rate": 3.62e-05,
      "loss": 0.01,
      "step": 2760
    },
    {
      "epoch": 0.277,
      "grad_norm": 0.11992328613996506,
      "learning_rate": 3.615e-05,
      "loss": 0.0107,
      "step": 2770
    },
    {
      "epoch": 0.278,
      "grad_norm": 0.10537979751825333,
      "learning_rate": 3.61e-05,
      "loss": 0.0101,
      "step": 2780
    },
    {
      "epoch": 0.279,
      "grad_norm": 0.06420467048883438,
      "learning_rate": 3.605e-05,
      "loss": 0.0088,
      "step": 2790
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.10813489556312561,
      "learning_rate": 3.6e-05,
      "loss": 0.0093,
      "step": 2800
    },
    {
      "epoch": 0.281,
      "grad_norm": 0.05735234543681145,
      "learning_rate": 3.595e-05,
      "loss": 0.0087,
      "step": 2810
    },
    {
      "epoch": 0.282,
      "grad_norm": 0.05712522938847542,
      "learning_rate": 3.59e-05,
      "loss": 0.0082,
      "step": 2820
    },
    {
      "epoch": 0.283,
      "grad_norm": 0.07710873335599899,
      "learning_rate": 3.585e-05,
      "loss": 0.0088,
      "step": 2830
    },
    {
      "epoch": 0.284,
      "grad_norm": 0.11007268726825714,
      "learning_rate": 3.58e-05,
      "loss": 0.0075,
      "step": 2840
    },
    {
      "epoch": 0.285,
      "grad_norm": 0.07825978100299835,
      "learning_rate": 3.575e-05,
      "loss": 0.0089,
      "step": 2850
    },
    {
      "epoch": 0.286,
      "grad_norm": 0.06950812041759491,
      "learning_rate": 3.57e-05,
      "loss": 0.0077,
      "step": 2860
    },
    {
      "epoch": 0.287,
      "grad_norm": 0.052572544664144516,
      "learning_rate": 3.565e-05,
      "loss": 0.0076,
      "step": 2870
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.06588669121265411,
      "learning_rate": 3.56e-05,
      "loss": 0.0073,
      "step": 2880
    },
    {
      "epoch": 0.289,
      "grad_norm": 0.052969031035900116,
      "learning_rate": 3.555e-05,
      "loss": 0.0066,
      "step": 2890
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.04204658418893814,
      "learning_rate": 3.55e-05,
      "loss": 0.0064,
      "step": 2900
    },
    {
      "epoch": 0.291,
      "grad_norm": 0.04765693470835686,
      "learning_rate": 3.545e-05,
      "loss": 0.0067,
      "step": 2910
    },
    {
      "epoch": 0.292,
      "grad_norm": 0.06796044856309891,
      "learning_rate": 3.54e-05,
      "loss": 0.0069,
      "step": 2920
    },
    {
      "epoch": 0.293,
      "grad_norm": 0.12173280119895935,
      "learning_rate": 3.535e-05,
      "loss": 0.0126,
      "step": 2930
    },
    {
      "epoch": 0.294,
      "grad_norm": 0.09393921494483948,
      "learning_rate": 3.53e-05,
      "loss": 0.0096,
      "step": 2940
    },
    {
      "epoch": 0.295,
      "grad_norm": 0.08246493339538574,
      "learning_rate": 3.525e-05,
      "loss": 0.0084,
      "step": 2950
    },
    {
      "epoch": 0.296,
      "grad_norm": 0.04482726752758026,
      "learning_rate": 3.52e-05,
      "loss": 0.0073,
      "step": 2960
    },
    {
      "epoch": 0.297,
      "grad_norm": 0.1147686317563057,
      "learning_rate": 3.515e-05,
      "loss": 0.0094,
      "step": 2970
    },
    {
      "epoch": 0.298,
      "grad_norm": 0.09143181890249252,
      "learning_rate": 3.51e-05,
      "loss": 0.0075,
      "step": 2980
    },
    {
      "epoch": 0.299,
      "grad_norm": 0.05911434814333916,
      "learning_rate": 3.505e-05,
      "loss": 0.0075,
      "step": 2990
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.04372965916991234,
      "learning_rate": 3.5e-05,
      "loss": 0.0072,
      "step": 3000
    },
    {
      "epoch": 0.301,
      "grad_norm": 0.05518479272723198,
      "learning_rate": 3.495e-05,
      "loss": 0.0068,
      "step": 3010
    },
    {
      "epoch": 0.302,
      "grad_norm": 0.04555105045437813,
      "learning_rate": 3.49e-05,
      "loss": 0.0064,
      "step": 3020
    },
    {
      "epoch": 0.303,
      "grad_norm": 0.03831150382757187,
      "learning_rate": 3.485e-05,
      "loss": 0.007,
      "step": 3030
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.04596749320626259,
      "learning_rate": 3.48e-05,
      "loss": 0.0065,
      "step": 3040
    },
    {
      "epoch": 0.305,
      "grad_norm": 0.07694078236818314,
      "learning_rate": 3.475e-05,
      "loss": 0.0059,
      "step": 3050
    },
    {
      "epoch": 0.306,
      "grad_norm": 0.12307348102331161,
      "learning_rate": 3.4699999999999996e-05,
      "loss": 0.0099,
      "step": 3060
    },
    {
      "epoch": 0.307,
      "grad_norm": 0.059611763805150986,
      "learning_rate": 3.465e-05,
      "loss": 0.0067,
      "step": 3070
    },
    {
      "epoch": 0.308,
      "grad_norm": 0.07357806712388992,
      "learning_rate": 3.46e-05,
      "loss": 0.007,
      "step": 3080
    },
    {
      "epoch": 0.309,
      "grad_norm": 0.060446444898843765,
      "learning_rate": 3.455e-05,
      "loss": 0.0063,
      "step": 3090
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.05178246274590492,
      "learning_rate": 3.45e-05,
      "loss": 0.0064,
      "step": 3100
    },
    {
      "epoch": 0.311,
      "grad_norm": 0.4560135006904602,
      "learning_rate": 3.445e-05,
      "loss": 0.0076,
      "step": 3110
    },
    {
      "epoch": 0.312,
      "grad_norm": 0.10910088568925858,
      "learning_rate": 3.4399999999999996e-05,
      "loss": 0.0082,
      "step": 3120
    },
    {
      "epoch": 0.313,
      "grad_norm": 0.05087321624159813,
      "learning_rate": 3.435e-05,
      "loss": 0.0061,
      "step": 3130
    },
    {
      "epoch": 0.314,
      "grad_norm": 0.055152568966150284,
      "learning_rate": 3.430000000000001e-05,
      "loss": 0.0065,
      "step": 3140
    },
    {
      "epoch": 0.315,
      "grad_norm": 0.48375555872917175,
      "learning_rate": 3.4250000000000006e-05,
      "loss": 0.0081,
      "step": 3150
    },
    {
      "epoch": 0.316,
      "grad_norm": 0.12223263829946518,
      "learning_rate": 3.4200000000000005e-05,
      "loss": 0.0098,
      "step": 3160
    },
    {
      "epoch": 0.317,
      "grad_norm": 0.15452681481838226,
      "learning_rate": 3.415e-05,
      "loss": 0.0087,
      "step": 3170
    },
    {
      "epoch": 0.318,
      "grad_norm": 0.06133843585848808,
      "learning_rate": 3.41e-05,
      "loss": 0.0087,
      "step": 3180
    },
    {
      "epoch": 0.319,
      "grad_norm": 0.04037950187921524,
      "learning_rate": 3.405e-05,
      "loss": 0.0062,
      "step": 3190
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.05707933381199837,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.0061,
      "step": 3200
    },
    {
      "epoch": 0.321,
      "grad_norm": 0.04124099016189575,
      "learning_rate": 3.3950000000000005e-05,
      "loss": 0.0058,
      "step": 3210
    },
    {
      "epoch": 0.322,
      "grad_norm": 0.12988638877868652,
      "learning_rate": 3.3900000000000004e-05,
      "loss": 0.0062,
      "step": 3220
    },
    {
      "epoch": 0.323,
      "grad_norm": 0.04961306229233742,
      "learning_rate": 3.385e-05,
      "loss": 0.0057,
      "step": 3230
    },
    {
      "epoch": 0.324,
      "grad_norm": 0.05354069173336029,
      "learning_rate": 3.38e-05,
      "loss": 0.0062,
      "step": 3240
    },
    {
      "epoch": 0.325,
      "grad_norm": 0.04944461211562157,
      "learning_rate": 3.375000000000001e-05,
      "loss": 0.0061,
      "step": 3250
    },
    {
      "epoch": 0.326,
      "grad_norm": 0.03429180383682251,
      "learning_rate": 3.3700000000000006e-05,
      "loss": 0.0055,
      "step": 3260
    },
    {
      "epoch": 0.327,
      "grad_norm": 0.05271946266293526,
      "learning_rate": 3.3650000000000005e-05,
      "loss": 0.0054,
      "step": 3270
    },
    {
      "epoch": 0.328,
      "grad_norm": 0.03602539002895355,
      "learning_rate": 3.3600000000000004e-05,
      "loss": 0.0049,
      "step": 3280
    },
    {
      "epoch": 0.329,
      "grad_norm": 0.03325178474187851,
      "learning_rate": 3.355e-05,
      "loss": 0.0052,
      "step": 3290
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.03728079795837402,
      "learning_rate": 3.35e-05,
      "loss": 0.0057,
      "step": 3300
    },
    {
      "epoch": 0.331,
      "grad_norm": 0.053768668323755264,
      "learning_rate": 3.345000000000001e-05,
      "loss": 0.0055,
      "step": 3310
    },
    {
      "epoch": 0.332,
      "grad_norm": 0.054501548409461975,
      "learning_rate": 3.3400000000000005e-05,
      "loss": 0.0053,
      "step": 3320
    },
    {
      "epoch": 0.333,
      "grad_norm": 0.05519956722855568,
      "learning_rate": 3.3350000000000004e-05,
      "loss": 0.0053,
      "step": 3330
    },
    {
      "epoch": 0.334,
      "grad_norm": 0.05373954027891159,
      "learning_rate": 3.33e-05,
      "loss": 0.006,
      "step": 3340
    },
    {
      "epoch": 0.335,
      "grad_norm": 0.04272560030221939,
      "learning_rate": 3.325e-05,
      "loss": 0.0057,
      "step": 3350
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.047061894088983536,
      "learning_rate": 3.32e-05,
      "loss": 0.0048,
      "step": 3360
    },
    {
      "epoch": 0.337,
      "grad_norm": 0.032794494181871414,
      "learning_rate": 3.3150000000000006e-05,
      "loss": 0.0046,
      "step": 3370
    },
    {
      "epoch": 0.338,
      "grad_norm": 0.027148200199007988,
      "learning_rate": 3.3100000000000005e-05,
      "loss": 0.0046,
      "step": 3380
    },
    {
      "epoch": 0.339,
      "grad_norm": 0.035516317933797836,
      "learning_rate": 3.3050000000000004e-05,
      "loss": 0.0048,
      "step": 3390
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.046294230967760086,
      "learning_rate": 3.3e-05,
      "loss": 0.0063,
      "step": 3400
    },
    {
      "epoch": 0.341,
      "grad_norm": 0.07840899378061295,
      "learning_rate": 3.295e-05,
      "loss": 0.0047,
      "step": 3410
    },
    {
      "epoch": 0.342,
      "grad_norm": 0.04392802715301514,
      "learning_rate": 3.29e-05,
      "loss": 0.0048,
      "step": 3420
    },
    {
      "epoch": 0.343,
      "grad_norm": 0.04237942770123482,
      "learning_rate": 3.2850000000000006e-05,
      "loss": 0.0047,
      "step": 3430
    },
    {
      "epoch": 0.344,
      "grad_norm": 0.03379204496741295,
      "learning_rate": 3.2800000000000004e-05,
      "loss": 0.005,
      "step": 3440
    },
    {
      "epoch": 0.345,
      "grad_norm": 0.3130718469619751,
      "learning_rate": 3.275e-05,
      "loss": 0.0108,
      "step": 3450
    },
    {
      "epoch": 0.346,
      "grad_norm": 0.13712112605571747,
      "learning_rate": 3.27e-05,
      "loss": 0.0095,
      "step": 3460
    },
    {
      "epoch": 0.347,
      "grad_norm": 0.12321494519710541,
      "learning_rate": 3.265e-05,
      "loss": 0.0075,
      "step": 3470
    },
    {
      "epoch": 0.348,
      "grad_norm": 0.06602940708398819,
      "learning_rate": 3.26e-05,
      "loss": 0.0062,
      "step": 3480
    },
    {
      "epoch": 0.349,
      "grad_norm": 0.08250287175178528,
      "learning_rate": 3.2550000000000005e-05,
      "loss": 0.0053,
      "step": 3490
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.04638442397117615,
      "learning_rate": 3.2500000000000004e-05,
      "loss": 0.0052,
      "step": 3500
    },
    {
      "epoch": 0.351,
      "grad_norm": 0.043373119086027145,
      "learning_rate": 3.245e-05,
      "loss": 0.0053,
      "step": 3510
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.04895636439323425,
      "learning_rate": 3.24e-05,
      "loss": 0.0043,
      "step": 3520
    },
    {
      "epoch": 0.353,
      "grad_norm": 0.04256746545433998,
      "learning_rate": 3.235e-05,
      "loss": 0.0044,
      "step": 3530
    },
    {
      "epoch": 0.354,
      "grad_norm": 0.02967280149459839,
      "learning_rate": 3.2300000000000006e-05,
      "loss": 0.0046,
      "step": 3540
    },
    {
      "epoch": 0.355,
      "grad_norm": 0.02590947411954403,
      "learning_rate": 3.2250000000000005e-05,
      "loss": 0.0044,
      "step": 3550
    },
    {
      "epoch": 0.356,
      "grad_norm": 0.026240160688757896,
      "learning_rate": 3.2200000000000003e-05,
      "loss": 0.0041,
      "step": 3560
    },
    {
      "epoch": 0.357,
      "grad_norm": 0.048163361847400665,
      "learning_rate": 3.215e-05,
      "loss": 0.0048,
      "step": 3570
    },
    {
      "epoch": 0.358,
      "grad_norm": 0.04310280829668045,
      "learning_rate": 3.21e-05,
      "loss": 0.0044,
      "step": 3580
    },
    {
      "epoch": 0.359,
      "grad_norm": 0.027477843686938286,
      "learning_rate": 3.205e-05,
      "loss": 0.004,
      "step": 3590
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.02773194946348667,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.0046,
      "step": 3600
    },
    {
      "epoch": 0.361,
      "grad_norm": 0.027110638096928596,
      "learning_rate": 3.1950000000000004e-05,
      "loss": 0.004,
      "step": 3610
    },
    {
      "epoch": 0.362,
      "grad_norm": 0.04346521571278572,
      "learning_rate": 3.19e-05,
      "loss": 0.0039,
      "step": 3620
    },
    {
      "epoch": 0.363,
      "grad_norm": 0.024588119238615036,
      "learning_rate": 3.185e-05,
      "loss": 0.0041,
      "step": 3630
    },
    {
      "epoch": 0.364,
      "grad_norm": 0.03631160408258438,
      "learning_rate": 3.18e-05,
      "loss": 0.0039,
      "step": 3640
    },
    {
      "epoch": 0.365,
      "grad_norm": 0.028497323393821716,
      "learning_rate": 3.175e-05,
      "loss": 0.0045,
      "step": 3650
    },
    {
      "epoch": 0.366,
      "grad_norm": 0.06324070692062378,
      "learning_rate": 3.1700000000000005e-05,
      "loss": 0.004,
      "step": 3660
    },
    {
      "epoch": 0.367,
      "grad_norm": 0.22121182084083557,
      "learning_rate": 3.1650000000000004e-05,
      "loss": 0.0078,
      "step": 3670
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.130402073264122,
      "learning_rate": 3.16e-05,
      "loss": 0.0068,
      "step": 3680
    },
    {
      "epoch": 0.369,
      "grad_norm": 0.11193361133337021,
      "learning_rate": 3.155e-05,
      "loss": 0.0052,
      "step": 3690
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.043261051177978516,
      "learning_rate": 3.15e-05,
      "loss": 0.0046,
      "step": 3700
    },
    {
      "epoch": 0.371,
      "grad_norm": 0.05132100731134415,
      "learning_rate": 3.145e-05,
      "loss": 0.0041,
      "step": 3710
    },
    {
      "epoch": 0.372,
      "grad_norm": 0.03498228266835213,
      "learning_rate": 3.1400000000000004e-05,
      "loss": 0.0041,
      "step": 3720
    },
    {
      "epoch": 0.373,
      "grad_norm": 0.03594733029603958,
      "learning_rate": 3.135e-05,
      "loss": 0.0037,
      "step": 3730
    },
    {
      "epoch": 0.374,
      "grad_norm": 0.05542735382914543,
      "learning_rate": 3.13e-05,
      "loss": 0.0042,
      "step": 3740
    },
    {
      "epoch": 0.375,
      "grad_norm": 0.03302931785583496,
      "learning_rate": 3.125e-05,
      "loss": 0.0048,
      "step": 3750
    },
    {
      "epoch": 0.376,
      "grad_norm": 0.028392167761921883,
      "learning_rate": 3.12e-05,
      "loss": 0.0042,
      "step": 3760
    },
    {
      "epoch": 0.377,
      "grad_norm": 0.05274713411927223,
      "learning_rate": 3.115e-05,
      "loss": 0.0045,
      "step": 3770
    },
    {
      "epoch": 0.378,
      "grad_norm": 0.024890929460525513,
      "learning_rate": 3.1100000000000004e-05,
      "loss": 0.0038,
      "step": 3780
    },
    {
      "epoch": 0.379,
      "grad_norm": 0.02797631174325943,
      "learning_rate": 3.105e-05,
      "loss": 0.0036,
      "step": 3790
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.033390454947948456,
      "learning_rate": 3.1e-05,
      "loss": 0.0039,
      "step": 3800
    },
    {
      "epoch": 0.381,
      "grad_norm": 0.024741416797041893,
      "learning_rate": 3.095e-05,
      "loss": 0.0042,
      "step": 3810
    },
    {
      "epoch": 0.382,
      "grad_norm": 0.05398337543010712,
      "learning_rate": 3.09e-05,
      "loss": 0.0049,
      "step": 3820
    },
    {
      "epoch": 0.383,
      "grad_norm": 0.027936646714806557,
      "learning_rate": 3.0850000000000004e-05,
      "loss": 0.0034,
      "step": 3830
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.02413495071232319,
      "learning_rate": 3.08e-05,
      "loss": 0.0036,
      "step": 3840
    },
    {
      "epoch": 0.385,
      "grad_norm": 0.037689995020627975,
      "learning_rate": 3.075e-05,
      "loss": 0.0033,
      "step": 3850
    },
    {
      "epoch": 0.386,
      "grad_norm": 0.028174949809908867,
      "learning_rate": 3.07e-05,
      "loss": 0.0036,
      "step": 3860
    },
    {
      "epoch": 0.387,
      "grad_norm": 0.064354807138443,
      "learning_rate": 3.065e-05,
      "loss": 0.0037,
      "step": 3870
    },
    {
      "epoch": 0.388,
      "grad_norm": 0.028341595083475113,
      "learning_rate": 3.06e-05,
      "loss": 0.0034,
      "step": 3880
    },
    {
      "epoch": 0.389,
      "grad_norm": 0.06142325699329376,
      "learning_rate": 3.0550000000000004e-05,
      "loss": 0.004,
      "step": 3890
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.03553822636604309,
      "learning_rate": 3.05e-05,
      "loss": 0.004,
      "step": 3900
    },
    {
      "epoch": 0.391,
      "grad_norm": 0.025645367801189423,
      "learning_rate": 3.045e-05,
      "loss": 0.0032,
      "step": 3910
    },
    {
      "epoch": 0.392,
      "grad_norm": 0.053947921842336655,
      "learning_rate": 3.04e-05,
      "loss": 0.0042,
      "step": 3920
    },
    {
      "epoch": 0.393,
      "grad_norm": 0.040126167237758636,
      "learning_rate": 3.035e-05,
      "loss": 0.0038,
      "step": 3930
    },
    {
      "epoch": 0.394,
      "grad_norm": 0.02956206165254116,
      "learning_rate": 3.03e-05,
      "loss": 0.0035,
      "step": 3940
    },
    {
      "epoch": 0.395,
      "grad_norm": 0.11952024698257446,
      "learning_rate": 3.025e-05,
      "loss": 0.0057,
      "step": 3950
    },
    {
      "epoch": 0.396,
      "grad_norm": 0.04155363142490387,
      "learning_rate": 3.02e-05,
      "loss": 0.0034,
      "step": 3960
    },
    {
      "epoch": 0.397,
      "grad_norm": 0.03884551301598549,
      "learning_rate": 3.015e-05,
      "loss": 0.0037,
      "step": 3970
    },
    {
      "epoch": 0.398,
      "grad_norm": 0.033869728446006775,
      "learning_rate": 3.01e-05,
      "loss": 0.0033,
      "step": 3980
    },
    {
      "epoch": 0.399,
      "grad_norm": 0.027508044615387917,
      "learning_rate": 3.0050000000000002e-05,
      "loss": 0.0038,
      "step": 3990
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.019838711246848106,
      "learning_rate": 3e-05,
      "loss": 0.0034,
      "step": 4000
    },
    {
      "epoch": 0.401,
      "grad_norm": 0.042124535888433456,
      "learning_rate": 2.995e-05,
      "loss": 0.003,
      "step": 4010
    },
    {
      "epoch": 0.402,
      "grad_norm": 0.03583139553666115,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 0.0029,
      "step": 4020
    },
    {
      "epoch": 0.403,
      "grad_norm": 0.024187223985791206,
      "learning_rate": 2.985e-05,
      "loss": 0.0031,
      "step": 4030
    },
    {
      "epoch": 0.404,
      "grad_norm": 0.02509123831987381,
      "learning_rate": 2.98e-05,
      "loss": 0.003,
      "step": 4040
    },
    {
      "epoch": 0.405,
      "grad_norm": 0.015798581764101982,
      "learning_rate": 2.975e-05,
      "loss": 0.003,
      "step": 4050
    },
    {
      "epoch": 0.406,
      "grad_norm": 0.01964486949145794,
      "learning_rate": 2.97e-05,
      "loss": 0.0032,
      "step": 4060
    },
    {
      "epoch": 0.407,
      "grad_norm": 0.025820232927799225,
      "learning_rate": 2.965e-05,
      "loss": 0.0032,
      "step": 4070
    },
    {
      "epoch": 0.408,
      "grad_norm": 0.03453589975833893,
      "learning_rate": 2.96e-05,
      "loss": 0.003,
      "step": 4080
    },
    {
      "epoch": 0.409,
      "grad_norm": 0.022311529144644737,
      "learning_rate": 2.955e-05,
      "loss": 0.0025,
      "step": 4090
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.02296466939151287,
      "learning_rate": 2.95e-05,
      "loss": 0.003,
      "step": 4100
    },
    {
      "epoch": 0.411,
      "grad_norm": 0.022816313430666924,
      "learning_rate": 2.945e-05,
      "loss": 0.0032,
      "step": 4110
    },
    {
      "epoch": 0.412,
      "grad_norm": 0.021030904725193977,
      "learning_rate": 2.94e-05,
      "loss": 0.003,
      "step": 4120
    },
    {
      "epoch": 0.413,
      "grad_norm": 0.02336346171796322,
      "learning_rate": 2.935e-05,
      "loss": 0.0028,
      "step": 4130
    },
    {
      "epoch": 0.414,
      "grad_norm": 0.019582638517022133,
      "learning_rate": 2.93e-05,
      "loss": 0.0027,
      "step": 4140
    },
    {
      "epoch": 0.415,
      "grad_norm": 0.031429585069417953,
      "learning_rate": 2.925e-05,
      "loss": 0.0027,
      "step": 4150
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.027804825454950333,
      "learning_rate": 2.9199999999999998e-05,
      "loss": 0.0027,
      "step": 4160
    },
    {
      "epoch": 0.417,
      "grad_norm": 0.022006656974554062,
      "learning_rate": 2.915e-05,
      "loss": 0.003,
      "step": 4170
    },
    {
      "epoch": 0.418,
      "grad_norm": 0.052478939294815063,
      "learning_rate": 2.91e-05,
      "loss": 0.0044,
      "step": 4180
    },
    {
      "epoch": 0.419,
      "grad_norm": 0.03854925185441971,
      "learning_rate": 2.9049999999999998e-05,
      "loss": 0.0036,
      "step": 4190
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.02749469131231308,
      "learning_rate": 2.9e-05,
      "loss": 0.0033,
      "step": 4200
    },
    {
      "epoch": 0.421,
      "grad_norm": 0.01697971485555172,
      "learning_rate": 2.895e-05,
      "loss": 0.0032,
      "step": 4210
    },
    {
      "epoch": 0.422,
      "grad_norm": 0.05183997377753258,
      "learning_rate": 2.8899999999999998e-05,
      "loss": 0.0029,
      "step": 4220
    },
    {
      "epoch": 0.423,
      "grad_norm": 0.030102815479040146,
      "learning_rate": 2.885e-05,
      "loss": 0.0028,
      "step": 4230
    },
    {
      "epoch": 0.424,
      "grad_norm": 0.027241216972470284,
      "learning_rate": 2.88e-05,
      "loss": 0.0033,
      "step": 4240
    },
    {
      "epoch": 0.425,
      "grad_norm": 0.01855759136378765,
      "learning_rate": 2.8749999999999997e-05,
      "loss": 0.0024,
      "step": 4250
    },
    {
      "epoch": 0.426,
      "grad_norm": 0.019300928339362144,
      "learning_rate": 2.87e-05,
      "loss": 0.0024,
      "step": 4260
    },
    {
      "epoch": 0.427,
      "grad_norm": 0.01639522798359394,
      "learning_rate": 2.865e-05,
      "loss": 0.0026,
      "step": 4270
    },
    {
      "epoch": 0.428,
      "grad_norm": 0.027084793895483017,
      "learning_rate": 2.86e-05,
      "loss": 0.0027,
      "step": 4280
    },
    {
      "epoch": 0.429,
      "grad_norm": 0.021206015720963478,
      "learning_rate": 2.855e-05,
      "loss": 0.0025,
      "step": 4290
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.0655827671289444,
      "learning_rate": 2.8499999999999998e-05,
      "loss": 0.0027,
      "step": 4300
    },
    {
      "epoch": 0.431,
      "grad_norm": 0.03779730945825577,
      "learning_rate": 2.845e-05,
      "loss": 0.0024,
      "step": 4310
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.045267749577760696,
      "learning_rate": 2.84e-05,
      "loss": 0.0031,
      "step": 4320
    },
    {
      "epoch": 0.433,
      "grad_norm": 0.017473919317126274,
      "learning_rate": 2.8349999999999998e-05,
      "loss": 0.0027,
      "step": 4330
    },
    {
      "epoch": 0.434,
      "grad_norm": 0.019513197243213654,
      "learning_rate": 2.83e-05,
      "loss": 0.0024,
      "step": 4340
    },
    {
      "epoch": 0.435,
      "grad_norm": 0.01616765186190605,
      "learning_rate": 2.825e-05,
      "loss": 0.0025,
      "step": 4350
    },
    {
      "epoch": 0.436,
      "grad_norm": 0.02270474284887314,
      "learning_rate": 2.8199999999999998e-05,
      "loss": 0.0024,
      "step": 4360
    },
    {
      "epoch": 0.437,
      "grad_norm": 0.02363002672791481,
      "learning_rate": 2.815e-05,
      "loss": 0.0025,
      "step": 4370
    },
    {
      "epoch": 0.438,
      "grad_norm": 0.023898936808109283,
      "learning_rate": 2.8100000000000005e-05,
      "loss": 0.0023,
      "step": 4380
    },
    {
      "epoch": 0.439,
      "grad_norm": 0.01270793005824089,
      "learning_rate": 2.8050000000000004e-05,
      "loss": 0.0022,
      "step": 4390
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.03917006403207779,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.0029,
      "step": 4400
    },
    {
      "epoch": 0.441,
      "grad_norm": 0.028284449130296707,
      "learning_rate": 2.7950000000000005e-05,
      "loss": 0.0023,
      "step": 4410
    },
    {
      "epoch": 0.442,
      "grad_norm": 0.017493903636932373,
      "learning_rate": 2.7900000000000004e-05,
      "loss": 0.0022,
      "step": 4420
    },
    {
      "epoch": 0.443,
      "grad_norm": 0.03160572797060013,
      "learning_rate": 2.7850000000000003e-05,
      "loss": 0.0025,
      "step": 4430
    },
    {
      "epoch": 0.444,
      "grad_norm": 0.022926049306988716,
      "learning_rate": 2.7800000000000005e-05,
      "loss": 0.0024,
      "step": 4440
    },
    {
      "epoch": 0.445,
      "grad_norm": 0.032902974635362625,
      "learning_rate": 2.7750000000000004e-05,
      "loss": 0.0025,
      "step": 4450
    },
    {
      "epoch": 0.446,
      "grad_norm": 0.017781972885131836,
      "learning_rate": 2.7700000000000002e-05,
      "loss": 0.0022,
      "step": 4460
    },
    {
      "epoch": 0.447,
      "grad_norm": 0.1209416389465332,
      "learning_rate": 2.7650000000000005e-05,
      "loss": 0.0028,
      "step": 4470
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.03747338801622391,
      "learning_rate": 2.7600000000000003e-05,
      "loss": 0.0024,
      "step": 4480
    },
    {
      "epoch": 0.449,
      "grad_norm": 0.04210735112428665,
      "learning_rate": 2.7550000000000002e-05,
      "loss": 0.0024,
      "step": 4490
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.02306324429810047,
      "learning_rate": 2.7500000000000004e-05,
      "loss": 0.0022,
      "step": 4500
    },
    {
      "epoch": 0.451,
      "grad_norm": 0.027622856199741364,
      "learning_rate": 2.7450000000000003e-05,
      "loss": 0.0026,
      "step": 4510
    },
    {
      "epoch": 0.452,
      "grad_norm": 0.014202162623405457,
      "learning_rate": 2.7400000000000002e-05,
      "loss": 0.0022,
      "step": 4520
    },
    {
      "epoch": 0.453,
      "grad_norm": 0.051466915756464005,
      "learning_rate": 2.7350000000000004e-05,
      "loss": 0.0031,
      "step": 4530
    },
    {
      "epoch": 0.454,
      "grad_norm": 0.052768610417842865,
      "learning_rate": 2.7300000000000003e-05,
      "loss": 0.0028,
      "step": 4540
    },
    {
      "epoch": 0.455,
      "grad_norm": 0.02291076071560383,
      "learning_rate": 2.725e-05,
      "loss": 0.0023,
      "step": 4550
    },
    {
      "epoch": 0.456,
      "grad_norm": 0.027942989021539688,
      "learning_rate": 2.7200000000000004e-05,
      "loss": 0.0023,
      "step": 4560
    },
    {
      "epoch": 0.457,
      "grad_norm": 0.01529670413583517,
      "learning_rate": 2.7150000000000003e-05,
      "loss": 0.0024,
      "step": 4570
    },
    {
      "epoch": 0.458,
      "grad_norm": 0.02945224940776825,
      "learning_rate": 2.7100000000000005e-05,
      "loss": 0.0024,
      "step": 4580
    },
    {
      "epoch": 0.459,
      "grad_norm": 0.027197351679205894,
      "learning_rate": 2.7050000000000004e-05,
      "loss": 0.0024,
      "step": 4590
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.022022951394319534,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 0.0023,
      "step": 4600
    },
    {
      "epoch": 0.461,
      "grad_norm": 0.019739823415875435,
      "learning_rate": 2.6950000000000005e-05,
      "loss": 0.0022,
      "step": 4610
    },
    {
      "epoch": 0.462,
      "grad_norm": 0.06794995814561844,
      "learning_rate": 2.6900000000000003e-05,
      "loss": 0.0027,
      "step": 4620
    },
    {
      "epoch": 0.463,
      "grad_norm": 0.049228962510824203,
      "learning_rate": 2.6850000000000002e-05,
      "loss": 0.0026,
      "step": 4630
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.0241558700799942,
      "learning_rate": 2.6800000000000004e-05,
      "loss": 0.0021,
      "step": 4640
    },
    {
      "epoch": 0.465,
      "grad_norm": 0.024576248601078987,
      "learning_rate": 2.6750000000000003e-05,
      "loss": 0.0022,
      "step": 4650
    },
    {
      "epoch": 0.466,
      "grad_norm": 0.030337205156683922,
      "learning_rate": 2.6700000000000002e-05,
      "loss": 0.0024,
      "step": 4660
    },
    {
      "epoch": 0.467,
      "grad_norm": 0.015081087127327919,
      "learning_rate": 2.6650000000000004e-05,
      "loss": 0.002,
      "step": 4670
    },
    {
      "epoch": 0.468,
      "grad_norm": 0.026368912309408188,
      "learning_rate": 2.6600000000000003e-05,
      "loss": 0.0022,
      "step": 4680
    },
    {
      "epoch": 0.469,
      "grad_norm": 0.018447600305080414,
      "learning_rate": 2.655e-05,
      "loss": 0.0022,
      "step": 4690
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.018314722925424576,
      "learning_rate": 2.6500000000000004e-05,
      "loss": 0.0019,
      "step": 4700
    },
    {
      "epoch": 0.471,
      "grad_norm": 0.02361704409122467,
      "learning_rate": 2.6450000000000003e-05,
      "loss": 0.0023,
      "step": 4710
    },
    {
      "epoch": 0.472,
      "grad_norm": 0.02032247930765152,
      "learning_rate": 2.64e-05,
      "loss": 0.002,
      "step": 4720
    },
    {
      "epoch": 0.473,
      "grad_norm": 0.017889728769659996,
      "learning_rate": 2.6350000000000004e-05,
      "loss": 0.0019,
      "step": 4730
    },
    {
      "epoch": 0.474,
      "grad_norm": 0.01962173730134964,
      "learning_rate": 2.6300000000000002e-05,
      "loss": 0.0019,
      "step": 4740
    },
    {
      "epoch": 0.475,
      "grad_norm": 0.015778113156557083,
      "learning_rate": 2.625e-05,
      "loss": 0.0021,
      "step": 4750
    },
    {
      "epoch": 0.476,
      "grad_norm": 0.01894952729344368,
      "learning_rate": 2.6200000000000003e-05,
      "loss": 0.0023,
      "step": 4760
    },
    {
      "epoch": 0.477,
      "grad_norm": 0.0239462498575449,
      "learning_rate": 2.6150000000000002e-05,
      "loss": 0.002,
      "step": 4770
    },
    {
      "epoch": 0.478,
      "grad_norm": 0.025406278669834137,
      "learning_rate": 2.61e-05,
      "loss": 0.0021,
      "step": 4780
    },
    {
      "epoch": 0.479,
      "grad_norm": 0.01813661865890026,
      "learning_rate": 2.6050000000000003e-05,
      "loss": 0.0023,
      "step": 4790
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.01979188807308674,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.0019,
      "step": 4800
    },
    {
      "epoch": 0.481,
      "grad_norm": 0.02219184674322605,
      "learning_rate": 2.595e-05,
      "loss": 0.002,
      "step": 4810
    },
    {
      "epoch": 0.482,
      "grad_norm": 0.012867514044046402,
      "learning_rate": 2.5900000000000003e-05,
      "loss": 0.0017,
      "step": 4820
    },
    {
      "epoch": 0.483,
      "grad_norm": 0.014178570359945297,
      "learning_rate": 2.585e-05,
      "loss": 0.0023,
      "step": 4830
    },
    {
      "epoch": 0.484,
      "grad_norm": 0.013582895509898663,
      "learning_rate": 2.58e-05,
      "loss": 0.002,
      "step": 4840
    },
    {
      "epoch": 0.485,
      "grad_norm": 0.02718137763440609,
      "learning_rate": 2.5750000000000002e-05,
      "loss": 0.0019,
      "step": 4850
    },
    {
      "epoch": 0.486,
      "grad_norm": 0.016559738665819168,
      "learning_rate": 2.57e-05,
      "loss": 0.0019,
      "step": 4860
    },
    {
      "epoch": 0.487,
      "grad_norm": 0.01447515282779932,
      "learning_rate": 2.5650000000000003e-05,
      "loss": 0.0017,
      "step": 4870
    },
    {
      "epoch": 0.488,
      "grad_norm": 0.010251802392303944,
      "learning_rate": 2.5600000000000002e-05,
      "loss": 0.0017,
      "step": 4880
    },
    {
      "epoch": 0.489,
      "grad_norm": 0.021138856187462807,
      "learning_rate": 2.555e-05,
      "loss": 0.0023,
      "step": 4890
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.026664163917303085,
      "learning_rate": 2.5500000000000003e-05,
      "loss": 0.0021,
      "step": 4900
    },
    {
      "epoch": 0.491,
      "grad_norm": 0.012794408947229385,
      "learning_rate": 2.5450000000000002e-05,
      "loss": 0.0017,
      "step": 4910
    },
    {
      "epoch": 0.492,
      "grad_norm": 0.013725240714848042,
      "learning_rate": 2.54e-05,
      "loss": 0.0018,
      "step": 4920
    },
    {
      "epoch": 0.493,
      "grad_norm": 0.01432815007865429,
      "learning_rate": 2.5350000000000003e-05,
      "loss": 0.0017,
      "step": 4930
    },
    {
      "epoch": 0.494,
      "grad_norm": 0.014761424623429775,
      "learning_rate": 2.5300000000000002e-05,
      "loss": 0.0019,
      "step": 4940
    },
    {
      "epoch": 0.495,
      "grad_norm": 0.06742983311414719,
      "learning_rate": 2.525e-05,
      "loss": 0.0025,
      "step": 4950
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.024261673912405968,
      "learning_rate": 2.5200000000000003e-05,
      "loss": 0.0021,
      "step": 4960
    },
    {
      "epoch": 0.497,
      "grad_norm": 0.02116272784769535,
      "learning_rate": 2.515e-05,
      "loss": 0.002,
      "step": 4970
    },
    {
      "epoch": 0.498,
      "grad_norm": 0.014996570535004139,
      "learning_rate": 2.51e-05,
      "loss": 0.0017,
      "step": 4980
    },
    {
      "epoch": 0.499,
      "grad_norm": 0.014970551244914532,
      "learning_rate": 2.5050000000000002e-05,
      "loss": 0.002,
      "step": 4990
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.01756688393652439,
      "learning_rate": 2.5e-05,
      "loss": 0.002,
      "step": 5000
    },
    {
      "epoch": 0.501,
      "grad_norm": 0.012683290056884289,
      "learning_rate": 2.495e-05,
      "loss": 0.0016,
      "step": 5010
    },
    {
      "epoch": 0.502,
      "grad_norm": 0.011495651677250862,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.0016,
      "step": 5020
    },
    {
      "epoch": 0.503,
      "grad_norm": 0.014306634664535522,
      "learning_rate": 2.485e-05,
      "loss": 0.0018,
      "step": 5030
    },
    {
      "epoch": 0.504,
      "grad_norm": 0.02241896465420723,
      "learning_rate": 2.48e-05,
      "loss": 0.0021,
      "step": 5040
    },
    {
      "epoch": 0.505,
      "grad_norm": 0.017740361392498016,
      "learning_rate": 2.4750000000000002e-05,
      "loss": 0.0016,
      "step": 5050
    },
    {
      "epoch": 0.506,
      "grad_norm": 0.013199679553508759,
      "learning_rate": 2.47e-05,
      "loss": 0.0015,
      "step": 5060
    },
    {
      "epoch": 0.507,
      "grad_norm": 0.057298243045806885,
      "learning_rate": 2.465e-05,
      "loss": 0.0019,
      "step": 5070
    },
    {
      "epoch": 0.508,
      "grad_norm": 0.03238265961408615,
      "learning_rate": 2.46e-05,
      "loss": 0.0026,
      "step": 5080
    },
    {
      "epoch": 0.509,
      "grad_norm": 0.04820936918258667,
      "learning_rate": 2.455e-05,
      "loss": 0.0027,
      "step": 5090
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.022526515647768974,
      "learning_rate": 2.45e-05,
      "loss": 0.0018,
      "step": 5100
    },
    {
      "epoch": 0.511,
      "grad_norm": 0.1899888962507248,
      "learning_rate": 2.445e-05,
      "loss": 0.0026,
      "step": 5110
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.05366889387369156,
      "learning_rate": 2.44e-05,
      "loss": 0.003,
      "step": 5120
    },
    {
      "epoch": 0.513,
      "grad_norm": 0.028939131647348404,
      "learning_rate": 2.435e-05,
      "loss": 0.0021,
      "step": 5130
    },
    {
      "epoch": 0.514,
      "grad_norm": 0.023352844640612602,
      "learning_rate": 2.43e-05,
      "loss": 0.0019,
      "step": 5140
    },
    {
      "epoch": 0.515,
      "grad_norm": 0.015283104963600636,
      "learning_rate": 2.425e-05,
      "loss": 0.0017,
      "step": 5150
    },
    {
      "epoch": 0.516,
      "grad_norm": 0.0149134686216712,
      "learning_rate": 2.4200000000000002e-05,
      "loss": 0.0016,
      "step": 5160
    },
    {
      "epoch": 0.517,
      "grad_norm": 0.01739874854683876,
      "learning_rate": 2.415e-05,
      "loss": 0.0021,
      "step": 5170
    },
    {
      "epoch": 0.518,
      "grad_norm": 0.012562318705022335,
      "learning_rate": 2.41e-05,
      "loss": 0.0016,
      "step": 5180
    },
    {
      "epoch": 0.519,
      "grad_norm": 0.01181173324584961,
      "learning_rate": 2.4050000000000002e-05,
      "loss": 0.0017,
      "step": 5190
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.0216183140873909,
      "learning_rate": 2.4e-05,
      "loss": 0.0017,
      "step": 5200
    },
    {
      "epoch": 0.521,
      "grad_norm": 0.014552557840943336,
      "learning_rate": 2.395e-05,
      "loss": 0.0017,
      "step": 5210
    },
    {
      "epoch": 0.522,
      "grad_norm": 0.013402258977293968,
      "learning_rate": 2.39e-05,
      "loss": 0.0015,
      "step": 5220
    },
    {
      "epoch": 0.523,
      "grad_norm": 0.017692307010293007,
      "learning_rate": 2.385e-05,
      "loss": 0.0017,
      "step": 5230
    },
    {
      "epoch": 0.524,
      "grad_norm": 0.007425515912473202,
      "learning_rate": 2.38e-05,
      "loss": 0.0015,
      "step": 5240
    },
    {
      "epoch": 0.525,
      "grad_norm": 0.010397032834589481,
      "learning_rate": 2.375e-05,
      "loss": 0.0014,
      "step": 5250
    },
    {
      "epoch": 0.526,
      "grad_norm": 0.013170558027923107,
      "learning_rate": 2.37e-05,
      "loss": 0.0017,
      "step": 5260
    },
    {
      "epoch": 0.527,
      "grad_norm": 0.47324055433273315,
      "learning_rate": 2.365e-05,
      "loss": 0.0037,
      "step": 5270
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.06395496428012848,
      "learning_rate": 2.36e-05,
      "loss": 0.003,
      "step": 5280
    },
    {
      "epoch": 0.529,
      "grad_norm": 0.032293129712343216,
      "learning_rate": 2.355e-05,
      "loss": 0.0022,
      "step": 5290
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.021514760330319405,
      "learning_rate": 2.35e-05,
      "loss": 0.002,
      "step": 5300
    },
    {
      "epoch": 0.531,
      "grad_norm": 0.016594447195529938,
      "learning_rate": 2.345e-05,
      "loss": 0.002,
      "step": 5310
    },
    {
      "epoch": 0.532,
      "grad_norm": 0.020661164075136185,
      "learning_rate": 2.3400000000000003e-05,
      "loss": 0.0018,
      "step": 5320
    },
    {
      "epoch": 0.533,
      "grad_norm": 0.01472094189375639,
      "learning_rate": 2.3350000000000002e-05,
      "loss": 0.0022,
      "step": 5330
    },
    {
      "epoch": 0.534,
      "grad_norm": 0.014501375146210194,
      "learning_rate": 2.3300000000000004e-05,
      "loss": 0.0017,
      "step": 5340
    },
    {
      "epoch": 0.535,
      "grad_norm": 0.01241264771670103,
      "learning_rate": 2.3250000000000003e-05,
      "loss": 0.0015,
      "step": 5350
    },
    {
      "epoch": 0.536,
      "grad_norm": 0.015589526854455471,
      "learning_rate": 2.32e-05,
      "loss": 0.0018,
      "step": 5360
    },
    {
      "epoch": 0.537,
      "grad_norm": 0.013468182645738125,
      "learning_rate": 2.3150000000000004e-05,
      "loss": 0.0018,
      "step": 5370
    },
    {
      "epoch": 0.538,
      "grad_norm": 0.015258733183145523,
      "learning_rate": 2.3100000000000002e-05,
      "loss": 0.0015,
      "step": 5380
    },
    {
      "epoch": 0.539,
      "grad_norm": 0.010932616889476776,
      "learning_rate": 2.305e-05,
      "loss": 0.0014,
      "step": 5390
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.0102313794195652,
      "learning_rate": 2.3000000000000003e-05,
      "loss": 0.0014,
      "step": 5400
    },
    {
      "epoch": 0.541,
      "grad_norm": 0.00674120569601655,
      "learning_rate": 2.2950000000000002e-05,
      "loss": 0.0014,
      "step": 5410
    },
    {
      "epoch": 0.542,
      "grad_norm": 0.015179513022303581,
      "learning_rate": 2.29e-05,
      "loss": 0.0014,
      "step": 5420
    },
    {
      "epoch": 0.543,
      "grad_norm": 0.03448422998189926,
      "learning_rate": 2.2850000000000003e-05,
      "loss": 0.0019,
      "step": 5430
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.028603358194231987,
      "learning_rate": 2.2800000000000002e-05,
      "loss": 0.0019,
      "step": 5440
    },
    {
      "epoch": 0.545,
      "grad_norm": 0.014372209087014198,
      "learning_rate": 2.275e-05,
      "loss": 0.0016,
      "step": 5450
    },
    {
      "epoch": 0.546,
      "grad_norm": 0.031532082706689835,
      "learning_rate": 2.2700000000000003e-05,
      "loss": 0.0017,
      "step": 5460
    },
    {
      "epoch": 0.547,
      "grad_norm": 0.018091056495904922,
      "learning_rate": 2.265e-05,
      "loss": 0.0016,
      "step": 5470
    },
    {
      "epoch": 0.548,
      "grad_norm": 0.014843069948256016,
      "learning_rate": 2.26e-05,
      "loss": 0.0015,
      "step": 5480
    },
    {
      "epoch": 0.549,
      "grad_norm": 0.011632148176431656,
      "learning_rate": 2.2550000000000003e-05,
      "loss": 0.0014,
      "step": 5490
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.009511668235063553,
      "learning_rate": 2.25e-05,
      "loss": 0.0014,
      "step": 5500
    },
    {
      "epoch": 0.551,
      "grad_norm": 0.007981637492775917,
      "learning_rate": 2.245e-05,
      "loss": 0.0014,
      "step": 5510
    },
    {
      "epoch": 0.552,
      "grad_norm": 0.021288806572556496,
      "learning_rate": 2.2400000000000002e-05,
      "loss": 0.0015,
      "step": 5520
    },
    {
      "epoch": 0.553,
      "grad_norm": 0.01468642894178629,
      "learning_rate": 2.235e-05,
      "loss": 0.0018,
      "step": 5530
    },
    {
      "epoch": 0.554,
      "grad_norm": 0.011532713659107685,
      "learning_rate": 2.23e-05,
      "loss": 0.0012,
      "step": 5540
    },
    {
      "epoch": 0.555,
      "grad_norm": 0.00889046210795641,
      "learning_rate": 2.2250000000000002e-05,
      "loss": 0.0011,
      "step": 5550
    },
    {
      "epoch": 0.556,
      "grad_norm": 0.01401284895837307,
      "learning_rate": 2.22e-05,
      "loss": 0.0014,
      "step": 5560
    },
    {
      "epoch": 0.557,
      "grad_norm": 0.012369327247142792,
      "learning_rate": 2.215e-05,
      "loss": 0.0015,
      "step": 5570
    },
    {
      "epoch": 0.558,
      "grad_norm": 0.015258446335792542,
      "learning_rate": 2.2100000000000002e-05,
      "loss": 0.0015,
      "step": 5580
    },
    {
      "epoch": 0.559,
      "grad_norm": 0.009015046060085297,
      "learning_rate": 2.205e-05,
      "loss": 0.0012,
      "step": 5590
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.011163819581270218,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.0012,
      "step": 5600
    },
    {
      "epoch": 0.561,
      "grad_norm": 0.016389524564146996,
      "learning_rate": 2.195e-05,
      "loss": 0.0016,
      "step": 5610
    },
    {
      "epoch": 0.562,
      "grad_norm": 0.01325678639113903,
      "learning_rate": 2.19e-05,
      "loss": 0.0013,
      "step": 5620
    },
    {
      "epoch": 0.563,
      "grad_norm": 0.017966121435165405,
      "learning_rate": 2.1850000000000003e-05,
      "loss": 0.0014,
      "step": 5630
    },
    {
      "epoch": 0.564,
      "grad_norm": 0.012039076536893845,
      "learning_rate": 2.18e-05,
      "loss": 0.0013,
      "step": 5640
    },
    {
      "epoch": 0.565,
      "grad_norm": 0.006665175314992666,
      "learning_rate": 2.175e-05,
      "loss": 0.0012,
      "step": 5650
    },
    {
      "epoch": 0.566,
      "grad_norm": 0.0105441864579916,
      "learning_rate": 2.1700000000000002e-05,
      "loss": 0.0014,
      "step": 5660
    },
    {
      "epoch": 0.567,
      "grad_norm": 0.007554101757705212,
      "learning_rate": 2.165e-05,
      "loss": 0.0011,
      "step": 5670
    },
    {
      "epoch": 0.568,
      "grad_norm": 0.009823901578783989,
      "learning_rate": 2.16e-05,
      "loss": 0.0013,
      "step": 5680
    },
    {
      "epoch": 0.569,
      "grad_norm": 0.01720455475151539,
      "learning_rate": 2.1550000000000002e-05,
      "loss": 0.0015,
      "step": 5690
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.01107338909059763,
      "learning_rate": 2.15e-05,
      "loss": 0.0012,
      "step": 5700
    },
    {
      "epoch": 0.571,
      "grad_norm": 0.01756761223077774,
      "learning_rate": 2.145e-05,
      "loss": 0.0014,
      "step": 5710
    },
    {
      "epoch": 0.572,
      "grad_norm": 0.022118983790278435,
      "learning_rate": 2.1400000000000002e-05,
      "loss": 0.0015,
      "step": 5720
    },
    {
      "epoch": 0.573,
      "grad_norm": 0.01616830937564373,
      "learning_rate": 2.135e-05,
      "loss": 0.0014,
      "step": 5730
    },
    {
      "epoch": 0.574,
      "grad_norm": 0.020481310784816742,
      "learning_rate": 2.13e-05,
      "loss": 0.0023,
      "step": 5740
    },
    {
      "epoch": 0.575,
      "grad_norm": 0.018176857382059097,
      "learning_rate": 2.125e-05,
      "loss": 0.0015,
      "step": 5750
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.011317101307213306,
      "learning_rate": 2.12e-05,
      "loss": 0.0012,
      "step": 5760
    },
    {
      "epoch": 0.577,
      "grad_norm": 0.028791502118110657,
      "learning_rate": 2.115e-05,
      "loss": 0.0014,
      "step": 5770
    },
    {
      "epoch": 0.578,
      "grad_norm": 0.013037024065852165,
      "learning_rate": 2.11e-05,
      "loss": 0.0013,
      "step": 5780
    },
    {
      "epoch": 0.579,
      "grad_norm": 0.021426070481538773,
      "learning_rate": 2.105e-05,
      "loss": 0.0015,
      "step": 5790
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.012033521197736263,
      "learning_rate": 2.1e-05,
      "loss": 0.0011,
      "step": 5800
    },
    {
      "epoch": 0.581,
      "grad_norm": 0.014337443746626377,
      "learning_rate": 2.095e-05,
      "loss": 0.0012,
      "step": 5810
    },
    {
      "epoch": 0.582,
      "grad_norm": 0.008603113703429699,
      "learning_rate": 2.09e-05,
      "loss": 0.0011,
      "step": 5820
    },
    {
      "epoch": 0.583,
      "grad_norm": 0.025418557226657867,
      "learning_rate": 2.085e-05,
      "loss": 0.0014,
      "step": 5830
    },
    {
      "epoch": 0.584,
      "grad_norm": 0.008621426299214363,
      "learning_rate": 2.08e-05,
      "loss": 0.0011,
      "step": 5840
    },
    {
      "epoch": 0.585,
      "grad_norm": 0.009969389997422695,
      "learning_rate": 2.075e-05,
      "loss": 0.0015,
      "step": 5850
    },
    {
      "epoch": 0.586,
      "grad_norm": 0.00997992418706417,
      "learning_rate": 2.07e-05,
      "loss": 0.0011,
      "step": 5860
    },
    {
      "epoch": 0.587,
      "grad_norm": 0.019949181005358696,
      "learning_rate": 2.065e-05,
      "loss": 0.001,
      "step": 5870
    },
    {
      "epoch": 0.588,
      "grad_norm": 0.009619793854653835,
      "learning_rate": 2.06e-05,
      "loss": 0.0011,
      "step": 5880
    },
    {
      "epoch": 0.589,
      "grad_norm": 0.007747489493340254,
      "learning_rate": 2.055e-05,
      "loss": 0.0012,
      "step": 5890
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.01052554789930582,
      "learning_rate": 2.05e-05,
      "loss": 0.0014,
      "step": 5900
    },
    {
      "epoch": 0.591,
      "grad_norm": 0.014904200099408627,
      "learning_rate": 2.045e-05,
      "loss": 0.0012,
      "step": 5910
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.00679561635479331,
      "learning_rate": 2.04e-05,
      "loss": 0.0011,
      "step": 5920
    },
    {
      "epoch": 0.593,
      "grad_norm": 0.006072670221328735,
      "learning_rate": 2.035e-05,
      "loss": 0.0011,
      "step": 5930
    },
    {
      "epoch": 0.594,
      "grad_norm": 0.014733157120645046,
      "learning_rate": 2.0300000000000002e-05,
      "loss": 0.0011,
      "step": 5940
    },
    {
      "epoch": 0.595,
      "grad_norm": 0.015511419624090195,
      "learning_rate": 2.025e-05,
      "loss": 0.0016,
      "step": 5950
    },
    {
      "epoch": 0.596,
      "grad_norm": 0.010620438493788242,
      "learning_rate": 2.0200000000000003e-05,
      "loss": 0.0012,
      "step": 5960
    },
    {
      "epoch": 0.597,
      "grad_norm": 0.0075794099830091,
      "learning_rate": 2.0150000000000002e-05,
      "loss": 0.0011,
      "step": 5970
    },
    {
      "epoch": 0.598,
      "grad_norm": 0.007882976904511452,
      "learning_rate": 2.01e-05,
      "loss": 0.0011,
      "step": 5980
    },
    {
      "epoch": 0.599,
      "grad_norm": 0.011548763141036034,
      "learning_rate": 2.0050000000000003e-05,
      "loss": 0.0013,
      "step": 5990
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.0084703853353858,
      "learning_rate": 2e-05,
      "loss": 0.0011,
      "step": 6000
    },
    {
      "epoch": 0.601,
      "grad_norm": 0.007603704463690519,
      "learning_rate": 1.995e-05,
      "loss": 0.001,
      "step": 6010
    },
    {
      "epoch": 0.602,
      "grad_norm": 0.008562711998820305,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 0.0012,
      "step": 6020
    },
    {
      "epoch": 0.603,
      "grad_norm": 0.007590813562273979,
      "learning_rate": 1.985e-05,
      "loss": 0.001,
      "step": 6030
    },
    {
      "epoch": 0.604,
      "grad_norm": 0.020342741161584854,
      "learning_rate": 1.9800000000000004e-05,
      "loss": 0.0017,
      "step": 6040
    },
    {
      "epoch": 0.605,
      "grad_norm": 0.16912633180618286,
      "learning_rate": 1.9750000000000002e-05,
      "loss": 0.0089,
      "step": 6050
    },
    {
      "epoch": 0.606,
      "grad_norm": 0.08793429285287857,
      "learning_rate": 1.97e-05,
      "loss": 0.0027,
      "step": 6060
    },
    {
      "epoch": 0.607,
      "grad_norm": 0.05196760594844818,
      "learning_rate": 1.9650000000000003e-05,
      "loss": 0.0022,
      "step": 6070
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.02118327096104622,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.0021,
      "step": 6080
    },
    {
      "epoch": 0.609,
      "grad_norm": 0.013289586640894413,
      "learning_rate": 1.955e-05,
      "loss": 0.0013,
      "step": 6090
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.012911707162857056,
      "learning_rate": 1.9500000000000003e-05,
      "loss": 0.0013,
      "step": 6100
    },
    {
      "epoch": 0.611,
      "grad_norm": 0.018663186579942703,
      "learning_rate": 1.9450000000000002e-05,
      "loss": 0.0012,
      "step": 6110
    },
    {
      "epoch": 0.612,
      "grad_norm": 0.010551884770393372,
      "learning_rate": 1.94e-05,
      "loss": 0.0012,
      "step": 6120
    },
    {
      "epoch": 0.613,
      "grad_norm": 0.015853077173233032,
      "learning_rate": 1.9350000000000003e-05,
      "loss": 0.0013,
      "step": 6130
    },
    {
      "epoch": 0.614,
      "grad_norm": 0.020374910905957222,
      "learning_rate": 1.93e-05,
      "loss": 0.001,
      "step": 6140
    },
    {
      "epoch": 0.615,
      "grad_norm": 0.015159848146140575,
      "learning_rate": 1.925e-05,
      "loss": 0.0013,
      "step": 6150
    },
    {
      "epoch": 0.616,
      "grad_norm": 0.007991676218807697,
      "learning_rate": 1.9200000000000003e-05,
      "loss": 0.0013,
      "step": 6160
    },
    {
      "epoch": 0.617,
      "grad_norm": 0.007849587127566338,
      "learning_rate": 1.915e-05,
      "loss": 0.0011,
      "step": 6170
    },
    {
      "epoch": 0.618,
      "grad_norm": 0.022048622369766235,
      "learning_rate": 1.91e-05,
      "loss": 0.001,
      "step": 6180
    },
    {
      "epoch": 0.619,
      "grad_norm": 0.021215343847870827,
      "learning_rate": 1.9050000000000002e-05,
      "loss": 0.0011,
      "step": 6190
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.012288344092667103,
      "learning_rate": 1.9e-05,
      "loss": 0.0012,
      "step": 6200
    },
    {
      "epoch": 0.621,
      "grad_norm": 0.020313331857323647,
      "learning_rate": 1.895e-05,
      "loss": 0.0011,
      "step": 6210
    },
    {
      "epoch": 0.622,
      "grad_norm": 0.008762447163462639,
      "learning_rate": 1.8900000000000002e-05,
      "loss": 0.001,
      "step": 6220
    },
    {
      "epoch": 0.623,
      "grad_norm": 0.0247616209089756,
      "learning_rate": 1.885e-05,
      "loss": 0.0011,
      "step": 6230
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.09021363407373428,
      "learning_rate": 1.88e-05,
      "loss": 0.0016,
      "step": 6240
    },
    {
      "epoch": 0.625,
      "grad_norm": 0.017945896834135056,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.0011,
      "step": 6250
    },
    {
      "epoch": 0.626,
      "grad_norm": 0.011303462088108063,
      "learning_rate": 1.87e-05,
      "loss": 0.0011,
      "step": 6260
    },
    {
      "epoch": 0.627,
      "grad_norm": 0.008381664752960205,
      "learning_rate": 1.865e-05,
      "loss": 0.0011,
      "step": 6270
    },
    {
      "epoch": 0.628,
      "grad_norm": 0.011003987863659859,
      "learning_rate": 1.86e-05,
      "loss": 0.0012,
      "step": 6280
    },
    {
      "epoch": 0.629,
      "grad_norm": 0.015965888276696205,
      "learning_rate": 1.855e-05,
      "loss": 0.001,
      "step": 6290
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.006507181562483311,
      "learning_rate": 1.85e-05,
      "loss": 0.0009,
      "step": 6300
    },
    {
      "epoch": 0.631,
      "grad_norm": 0.015577591024339199,
      "learning_rate": 1.845e-05,
      "loss": 0.001,
      "step": 6310
    },
    {
      "epoch": 0.632,
      "grad_norm": 0.006741558667272329,
      "learning_rate": 1.84e-05,
      "loss": 0.0011,
      "step": 6320
    },
    {
      "epoch": 0.633,
      "grad_norm": 0.016030525788664818,
      "learning_rate": 1.8350000000000002e-05,
      "loss": 0.001,
      "step": 6330
    },
    {
      "epoch": 0.634,
      "grad_norm": 0.010763168334960938,
      "learning_rate": 1.83e-05,
      "loss": 0.0011,
      "step": 6340
    },
    {
      "epoch": 0.635,
      "grad_norm": 0.017273874953389168,
      "learning_rate": 1.825e-05,
      "loss": 0.001,
      "step": 6350
    },
    {
      "epoch": 0.636,
      "grad_norm": 0.010964670218527317,
      "learning_rate": 1.8200000000000002e-05,
      "loss": 0.0011,
      "step": 6360
    },
    {
      "epoch": 0.637,
      "grad_norm": 0.00803497713059187,
      "learning_rate": 1.815e-05,
      "loss": 0.0009,
      "step": 6370
    },
    {
      "epoch": 0.638,
      "grad_norm": 0.007479315157979727,
      "learning_rate": 1.81e-05,
      "loss": 0.0014,
      "step": 6380
    },
    {
      "epoch": 0.639,
      "grad_norm": 0.010598058812320232,
      "learning_rate": 1.805e-05,
      "loss": 0.001,
      "step": 6390
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.009770036675035954,
      "learning_rate": 1.8e-05,
      "loss": 0.0009,
      "step": 6400
    },
    {
      "epoch": 0.641,
      "grad_norm": 0.011602561920881271,
      "learning_rate": 1.795e-05,
      "loss": 0.0008,
      "step": 6410
    },
    {
      "epoch": 0.642,
      "grad_norm": 0.0076597342267632484,
      "learning_rate": 1.79e-05,
      "loss": 0.0008,
      "step": 6420
    },
    {
      "epoch": 0.643,
      "grad_norm": 0.012248953804373741,
      "learning_rate": 1.785e-05,
      "loss": 0.0008,
      "step": 6430
    },
    {
      "epoch": 0.644,
      "grad_norm": 0.005626557394862175,
      "learning_rate": 1.78e-05,
      "loss": 0.0008,
      "step": 6440
    },
    {
      "epoch": 0.645,
      "grad_norm": 0.005482000298798084,
      "learning_rate": 1.775e-05,
      "loss": 0.0008,
      "step": 6450
    },
    {
      "epoch": 0.646,
      "grad_norm": 0.007456011138856411,
      "learning_rate": 1.77e-05,
      "loss": 0.0008,
      "step": 6460
    },
    {
      "epoch": 0.647,
      "grad_norm": 0.008909308351576328,
      "learning_rate": 1.765e-05,
      "loss": 0.0008,
      "step": 6470
    },
    {
      "epoch": 0.648,
      "grad_norm": 0.011135280132293701,
      "learning_rate": 1.76e-05,
      "loss": 0.0009,
      "step": 6480
    },
    {
      "epoch": 0.649,
      "grad_norm": 0.01595783233642578,
      "learning_rate": 1.755e-05,
      "loss": 0.001,
      "step": 6490
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.013902807608246803,
      "learning_rate": 1.75e-05,
      "loss": 0.0011,
      "step": 6500
    },
    {
      "epoch": 0.651,
      "grad_norm": 0.010244622826576233,
      "learning_rate": 1.745e-05,
      "loss": 0.0009,
      "step": 6510
    },
    {
      "epoch": 0.652,
      "grad_norm": 0.007476091384887695,
      "learning_rate": 1.74e-05,
      "loss": 0.0009,
      "step": 6520
    },
    {
      "epoch": 0.653,
      "grad_norm": 0.013044660910964012,
      "learning_rate": 1.7349999999999998e-05,
      "loss": 0.0009,
      "step": 6530
    },
    {
      "epoch": 0.654,
      "grad_norm": 0.004804369527846575,
      "learning_rate": 1.73e-05,
      "loss": 0.0009,
      "step": 6540
    },
    {
      "epoch": 0.655,
      "grad_norm": 0.006042002234607935,
      "learning_rate": 1.725e-05,
      "loss": 0.0008,
      "step": 6550
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.010785943828523159,
      "learning_rate": 1.7199999999999998e-05,
      "loss": 0.0009,
      "step": 6560
    },
    {
      "epoch": 0.657,
      "grad_norm": 0.011350172571837902,
      "learning_rate": 1.7150000000000004e-05,
      "loss": 0.0008,
      "step": 6570
    },
    {
      "epoch": 0.658,
      "grad_norm": 0.007638021372258663,
      "learning_rate": 1.7100000000000002e-05,
      "loss": 0.0009,
      "step": 6580
    },
    {
      "epoch": 0.659,
      "grad_norm": 0.005735939834266901,
      "learning_rate": 1.705e-05,
      "loss": 0.0009,
      "step": 6590
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.02717960625886917,
      "learning_rate": 1.7000000000000003e-05,
      "loss": 0.0011,
      "step": 6600
    },
    {
      "epoch": 0.661,
      "grad_norm": 0.006012643221765757,
      "learning_rate": 1.6950000000000002e-05,
      "loss": 0.0008,
      "step": 6610
    },
    {
      "epoch": 0.662,
      "grad_norm": 0.00599683728069067,
      "learning_rate": 1.69e-05,
      "loss": 0.0008,
      "step": 6620
    },
    {
      "epoch": 0.663,
      "grad_norm": 0.026952974498271942,
      "learning_rate": 1.6850000000000003e-05,
      "loss": 0.0008,
      "step": 6630
    },
    {
      "epoch": 0.664,
      "grad_norm": 0.008171536959707737,
      "learning_rate": 1.6800000000000002e-05,
      "loss": 0.0008,
      "step": 6640
    },
    {
      "epoch": 0.665,
      "grad_norm": 0.007446442265063524,
      "learning_rate": 1.675e-05,
      "loss": 0.0009,
      "step": 6650
    },
    {
      "epoch": 0.666,
      "grad_norm": 0.006456063129007816,
      "learning_rate": 1.6700000000000003e-05,
      "loss": 0.0008,
      "step": 6660
    },
    {
      "epoch": 0.667,
      "grad_norm": 0.008162173442542553,
      "learning_rate": 1.665e-05,
      "loss": 0.0007,
      "step": 6670
    },
    {
      "epoch": 0.668,
      "grad_norm": 0.004432919900864363,
      "learning_rate": 1.66e-05,
      "loss": 0.0008,
      "step": 6680
    },
    {
      "epoch": 0.669,
      "grad_norm": 0.007158307824283838,
      "learning_rate": 1.6550000000000002e-05,
      "loss": 0.0008,
      "step": 6690
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.003983801696449518,
      "learning_rate": 1.65e-05,
      "loss": 0.0007,
      "step": 6700
    },
    {
      "epoch": 0.671,
      "grad_norm": 0.005170087795704603,
      "learning_rate": 1.645e-05,
      "loss": 0.0008,
      "step": 6710
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.004729804117232561,
      "learning_rate": 1.6400000000000002e-05,
      "loss": 0.0008,
      "step": 6720
    },
    {
      "epoch": 0.673,
      "grad_norm": 0.010037174448370934,
      "learning_rate": 1.635e-05,
      "loss": 0.001,
      "step": 6730
    },
    {
      "epoch": 0.674,
      "grad_norm": 0.050949569791555405,
      "learning_rate": 1.63e-05,
      "loss": 0.0023,
      "step": 6740
    },
    {
      "epoch": 0.675,
      "grad_norm": 0.0323474146425724,
      "learning_rate": 1.6250000000000002e-05,
      "loss": 0.0017,
      "step": 6750
    },
    {
      "epoch": 0.676,
      "grad_norm": 0.027231359854340553,
      "learning_rate": 1.62e-05,
      "loss": 0.0021,
      "step": 6760
    },
    {
      "epoch": 0.677,
      "grad_norm": 0.01555855292826891,
      "learning_rate": 1.6150000000000003e-05,
      "loss": 0.0013,
      "step": 6770
    },
    {
      "epoch": 0.678,
      "grad_norm": 0.01804298162460327,
      "learning_rate": 1.6100000000000002e-05,
      "loss": 0.0011,
      "step": 6780
    },
    {
      "epoch": 0.679,
      "grad_norm": 0.011248771101236343,
      "learning_rate": 1.605e-05,
      "loss": 0.0011,
      "step": 6790
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.007389044389128685,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.0009,
      "step": 6800
    },
    {
      "epoch": 0.681,
      "grad_norm": 0.014606145210564137,
      "learning_rate": 1.595e-05,
      "loss": 0.0012,
      "step": 6810
    },
    {
      "epoch": 0.682,
      "grad_norm": 0.012476052157580853,
      "learning_rate": 1.59e-05,
      "loss": 0.0009,
      "step": 6820
    },
    {
      "epoch": 0.683,
      "grad_norm": 0.009272475726902485,
      "learning_rate": 1.5850000000000002e-05,
      "loss": 0.0009,
      "step": 6830
    },
    {
      "epoch": 0.684,
      "grad_norm": 0.011705187149345875,
      "learning_rate": 1.58e-05,
      "loss": 0.0009,
      "step": 6840
    },
    {
      "epoch": 0.685,
      "grad_norm": 0.01874556578695774,
      "learning_rate": 1.575e-05,
      "loss": 0.0011,
      "step": 6850
    },
    {
      "epoch": 0.686,
      "grad_norm": 0.01463324110955,
      "learning_rate": 1.5700000000000002e-05,
      "loss": 0.0009,
      "step": 6860
    },
    {
      "epoch": 0.687,
      "grad_norm": 0.012001392431557178,
      "learning_rate": 1.565e-05,
      "loss": 0.001,
      "step": 6870
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.009366356767714024,
      "learning_rate": 1.56e-05,
      "loss": 0.0008,
      "step": 6880
    },
    {
      "epoch": 0.689,
      "grad_norm": 0.010064000263810158,
      "learning_rate": 1.5550000000000002e-05,
      "loss": 0.0009,
      "step": 6890
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.016703909263014793,
      "learning_rate": 1.55e-05,
      "loss": 0.0009,
      "step": 6900
    },
    {
      "epoch": 0.691,
      "grad_norm": 0.0146669652312994,
      "learning_rate": 1.545e-05,
      "loss": 0.001,
      "step": 6910
    },
    {
      "epoch": 0.692,
      "grad_norm": 0.006643705535680056,
      "learning_rate": 1.54e-05,
      "loss": 0.0009,
      "step": 6920
    },
    {
      "epoch": 0.693,
      "grad_norm": 0.011501871049404144,
      "learning_rate": 1.535e-05,
      "loss": 0.0008,
      "step": 6930
    },
    {
      "epoch": 0.694,
      "grad_norm": 0.008170065470039845,
      "learning_rate": 1.53e-05,
      "loss": 0.0008,
      "step": 6940
    },
    {
      "epoch": 0.695,
      "grad_norm": 0.00737554719671607,
      "learning_rate": 1.525e-05,
      "loss": 0.0007,
      "step": 6950
    },
    {
      "epoch": 0.696,
      "grad_norm": 0.006846282631158829,
      "learning_rate": 1.52e-05,
      "loss": 0.0009,
      "step": 6960
    },
    {
      "epoch": 0.697,
      "grad_norm": 0.007784941233694553,
      "learning_rate": 1.515e-05,
      "loss": 0.0008,
      "step": 6970
    },
    {
      "epoch": 0.698,
      "grad_norm": 0.009864069521427155,
      "learning_rate": 1.51e-05,
      "loss": 0.0008,
      "step": 6980
    },
    {
      "epoch": 0.699,
      "grad_norm": 0.007372863125056028,
      "learning_rate": 1.505e-05,
      "loss": 0.0009,
      "step": 6990
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.006507135462015867,
      "learning_rate": 1.5e-05,
      "loss": 0.0008,
      "step": 7000
    },
    {
      "epoch": 0.701,
      "grad_norm": 0.03093353845179081,
      "learning_rate": 1.4950000000000001e-05,
      "loss": 0.0014,
      "step": 7010
    },
    {
      "epoch": 0.702,
      "grad_norm": 0.01417300570756197,
      "learning_rate": 1.49e-05,
      "loss": 0.001,
      "step": 7020
    },
    {
      "epoch": 0.703,
      "grad_norm": 0.010836401022970676,
      "learning_rate": 1.485e-05,
      "loss": 0.0012,
      "step": 7030
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.01000068336725235,
      "learning_rate": 1.48e-05,
      "loss": 0.001,
      "step": 7040
    },
    {
      "epoch": 0.705,
      "grad_norm": 0.008654952049255371,
      "learning_rate": 1.475e-05,
      "loss": 0.0009,
      "step": 7050
    },
    {
      "epoch": 0.706,
      "grad_norm": 0.010761331766843796,
      "learning_rate": 1.47e-05,
      "loss": 0.001,
      "step": 7060
    },
    {
      "epoch": 0.707,
      "grad_norm": 0.006188638508319855,
      "learning_rate": 1.465e-05,
      "loss": 0.0008,
      "step": 7070
    },
    {
      "epoch": 0.708,
      "grad_norm": 0.007858789525926113,
      "learning_rate": 1.4599999999999999e-05,
      "loss": 0.0008,
      "step": 7080
    },
    {
      "epoch": 0.709,
      "grad_norm": 0.02773350477218628,
      "learning_rate": 1.455e-05,
      "loss": 0.0014,
      "step": 7090
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.012381108477711678,
      "learning_rate": 1.45e-05,
      "loss": 0.0009,
      "step": 7100
    },
    {
      "epoch": 0.711,
      "grad_norm": 0.009256324730813503,
      "learning_rate": 1.4449999999999999e-05,
      "loss": 0.0008,
      "step": 7110
    },
    {
      "epoch": 0.712,
      "grad_norm": 0.007005748804658651,
      "learning_rate": 1.44e-05,
      "loss": 0.0009,
      "step": 7120
    },
    {
      "epoch": 0.713,
      "grad_norm": 0.0055755749344825745,
      "learning_rate": 1.435e-05,
      "loss": 0.0007,
      "step": 7130
    },
    {
      "epoch": 0.714,
      "grad_norm": 0.003967254888266325,
      "learning_rate": 1.43e-05,
      "loss": 0.0008,
      "step": 7140
    },
    {
      "epoch": 0.715,
      "grad_norm": 0.0079165268689394,
      "learning_rate": 1.4249999999999999e-05,
      "loss": 0.0011,
      "step": 7150
    },
    {
      "epoch": 0.716,
      "grad_norm": 0.004682580940425396,
      "learning_rate": 1.42e-05,
      "loss": 0.0007,
      "step": 7160
    },
    {
      "epoch": 0.717,
      "grad_norm": 0.008578700013458729,
      "learning_rate": 1.415e-05,
      "loss": 0.0011,
      "step": 7170
    },
    {
      "epoch": 0.718,
      "grad_norm": 0.006943961605429649,
      "learning_rate": 1.4099999999999999e-05,
      "loss": 0.0009,
      "step": 7180
    },
    {
      "epoch": 0.719,
      "grad_norm": 0.0072656250558793545,
      "learning_rate": 1.4050000000000003e-05,
      "loss": 0.0007,
      "step": 7190
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.005639955401420593,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.0007,
      "step": 7200
    },
    {
      "epoch": 0.721,
      "grad_norm": 0.005733838304877281,
      "learning_rate": 1.3950000000000002e-05,
      "loss": 0.0008,
      "step": 7210
    },
    {
      "epoch": 0.722,
      "grad_norm": 0.02654002234339714,
      "learning_rate": 1.3900000000000002e-05,
      "loss": 0.0008,
      "step": 7220
    },
    {
      "epoch": 0.723,
      "grad_norm": 0.007308628410100937,
      "learning_rate": 1.3850000000000001e-05,
      "loss": 0.0008,
      "step": 7230
    },
    {
      "epoch": 0.724,
      "grad_norm": 0.006939894054085016,
      "learning_rate": 1.3800000000000002e-05,
      "loss": 0.0007,
      "step": 7240
    },
    {
      "epoch": 0.725,
      "grad_norm": 0.03964811936020851,
      "learning_rate": 1.3750000000000002e-05,
      "loss": 0.0013,
      "step": 7250
    },
    {
      "epoch": 0.726,
      "grad_norm": 0.014138396829366684,
      "learning_rate": 1.3700000000000001e-05,
      "loss": 0.001,
      "step": 7260
    },
    {
      "epoch": 0.727,
      "grad_norm": 0.008445181883871555,
      "learning_rate": 1.3650000000000001e-05,
      "loss": 0.0008,
      "step": 7270
    },
    {
      "epoch": 0.728,
      "grad_norm": 0.01134855579584837,
      "learning_rate": 1.3600000000000002e-05,
      "loss": 0.0009,
      "step": 7280
    },
    {
      "epoch": 0.729,
      "grad_norm": 0.010982022620737553,
      "learning_rate": 1.3550000000000002e-05,
      "loss": 0.0015,
      "step": 7290
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.011698734015226364,
      "learning_rate": 1.3500000000000001e-05,
      "loss": 0.0008,
      "step": 7300
    },
    {
      "epoch": 0.731,
      "grad_norm": 0.006420729216188192,
      "learning_rate": 1.3450000000000002e-05,
      "loss": 0.0008,
      "step": 7310
    },
    {
      "epoch": 0.732,
      "grad_norm": 0.006088167428970337,
      "learning_rate": 1.3400000000000002e-05,
      "loss": 0.0008,
      "step": 7320
    },
    {
      "epoch": 0.733,
      "grad_norm": 0.0071141645312309265,
      "learning_rate": 1.3350000000000001e-05,
      "loss": 0.0012,
      "step": 7330
    },
    {
      "epoch": 0.734,
      "grad_norm": 0.004975921008735895,
      "learning_rate": 1.3300000000000001e-05,
      "loss": 0.0006,
      "step": 7340
    },
    {
      "epoch": 0.735,
      "grad_norm": 0.004499469883739948,
      "learning_rate": 1.3250000000000002e-05,
      "loss": 0.0007,
      "step": 7350
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.009738982655107975,
      "learning_rate": 1.32e-05,
      "loss": 0.001,
      "step": 7360
    },
    {
      "epoch": 0.737,
      "grad_norm": 0.006863337475806475,
      "learning_rate": 1.3150000000000001e-05,
      "loss": 0.001,
      "step": 7370
    },
    {
      "epoch": 0.738,
      "grad_norm": 0.008216536603868008,
      "learning_rate": 1.3100000000000002e-05,
      "loss": 0.0007,
      "step": 7380
    },
    {
      "epoch": 0.739,
      "grad_norm": 0.006803369149565697,
      "learning_rate": 1.305e-05,
      "loss": 0.0008,
      "step": 7390
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.00551017839461565,
      "learning_rate": 1.3000000000000001e-05,
      "loss": 0.0008,
      "step": 7400
    },
    {
      "epoch": 0.741,
      "grad_norm": 0.009463651105761528,
      "learning_rate": 1.2950000000000001e-05,
      "loss": 0.0008,
      "step": 7410
    },
    {
      "epoch": 0.742,
      "grad_norm": 0.01233983039855957,
      "learning_rate": 1.29e-05,
      "loss": 0.0019,
      "step": 7420
    },
    {
      "epoch": 0.743,
      "grad_norm": 0.008470877073705196,
      "learning_rate": 1.285e-05,
      "loss": 0.0009,
      "step": 7430
    },
    {
      "epoch": 0.744,
      "grad_norm": 0.007592742796987295,
      "learning_rate": 1.2800000000000001e-05,
      "loss": 0.0008,
      "step": 7440
    },
    {
      "epoch": 0.745,
      "grad_norm": 0.03596987947821617,
      "learning_rate": 1.2750000000000002e-05,
      "loss": 0.001,
      "step": 7450
    },
    {
      "epoch": 0.746,
      "grad_norm": 0.005849502049386501,
      "learning_rate": 1.27e-05,
      "loss": 0.0008,
      "step": 7460
    },
    {
      "epoch": 0.747,
      "grad_norm": 0.009035659022629261,
      "learning_rate": 1.2650000000000001e-05,
      "loss": 0.0007,
      "step": 7470
    },
    {
      "epoch": 0.748,
      "grad_norm": 0.010397679172456264,
      "learning_rate": 1.2600000000000001e-05,
      "loss": 0.0014,
      "step": 7480
    },
    {
      "epoch": 0.749,
      "grad_norm": 0.014514378271996975,
      "learning_rate": 1.255e-05,
      "loss": 0.0008,
      "step": 7490
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.004837281536310911,
      "learning_rate": 1.25e-05,
      "loss": 0.0006,
      "step": 7500
    },
    {
      "epoch": 0.751,
      "grad_norm": 0.007720770314335823,
      "learning_rate": 1.2450000000000001e-05,
      "loss": 0.0006,
      "step": 7510
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.012046804651618004,
      "learning_rate": 1.24e-05,
      "loss": 0.0011,
      "step": 7520
    },
    {
      "epoch": 0.753,
      "grad_norm": 0.01343387458473444,
      "learning_rate": 1.235e-05,
      "loss": 0.0007,
      "step": 7530
    },
    {
      "epoch": 0.754,
      "grad_norm": 0.00810600072145462,
      "learning_rate": 1.23e-05,
      "loss": 0.0007,
      "step": 7540
    },
    {
      "epoch": 0.755,
      "grad_norm": 0.00925883837044239,
      "learning_rate": 1.225e-05,
      "loss": 0.0006,
      "step": 7550
    },
    {
      "epoch": 0.756,
      "grad_norm": 0.01927885413169861,
      "learning_rate": 1.22e-05,
      "loss": 0.0014,
      "step": 7560
    },
    {
      "epoch": 0.757,
      "grad_norm": 0.010129665955901146,
      "learning_rate": 1.215e-05,
      "loss": 0.0006,
      "step": 7570
    },
    {
      "epoch": 0.758,
      "grad_norm": 0.007863885723054409,
      "learning_rate": 1.2100000000000001e-05,
      "loss": 0.0006,
      "step": 7580
    },
    {
      "epoch": 0.759,
      "grad_norm": 0.005500464700162411,
      "learning_rate": 1.205e-05,
      "loss": 0.0007,
      "step": 7590
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.0040563903748989105,
      "learning_rate": 1.2e-05,
      "loss": 0.0006,
      "step": 7600
    },
    {
      "epoch": 0.761,
      "grad_norm": 0.006361998151987791,
      "learning_rate": 1.195e-05,
      "loss": 0.0007,
      "step": 7610
    },
    {
      "epoch": 0.762,
      "grad_norm": 0.0136310625821352,
      "learning_rate": 1.19e-05,
      "loss": 0.0008,
      "step": 7620
    },
    {
      "epoch": 0.763,
      "grad_norm": 0.005384715739637613,
      "learning_rate": 1.185e-05,
      "loss": 0.0007,
      "step": 7630
    },
    {
      "epoch": 0.764,
      "grad_norm": 0.014707676135003567,
      "learning_rate": 1.18e-05,
      "loss": 0.0007,
      "step": 7640
    },
    {
      "epoch": 0.765,
      "grad_norm": 0.008092684671282768,
      "learning_rate": 1.175e-05,
      "loss": 0.0006,
      "step": 7650
    },
    {
      "epoch": 0.766,
      "grad_norm": 0.007185132242739201,
      "learning_rate": 1.1700000000000001e-05,
      "loss": 0.0006,
      "step": 7660
    },
    {
      "epoch": 0.767,
      "grad_norm": 0.005672789178788662,
      "learning_rate": 1.1650000000000002e-05,
      "loss": 0.0006,
      "step": 7670
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.05434956029057503,
      "learning_rate": 1.16e-05,
      "loss": 0.001,
      "step": 7680
    },
    {
      "epoch": 0.769,
      "grad_norm": 0.00933472067117691,
      "learning_rate": 1.1550000000000001e-05,
      "loss": 0.0007,
      "step": 7690
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.008684621192514896,
      "learning_rate": 1.1500000000000002e-05,
      "loss": 0.0006,
      "step": 7700
    },
    {
      "epoch": 0.771,
      "grad_norm": 0.03054739721119404,
      "learning_rate": 1.145e-05,
      "loss": 0.0006,
      "step": 7710
    },
    {
      "epoch": 0.772,
      "grad_norm": 0.005998207256197929,
      "learning_rate": 1.1400000000000001e-05,
      "loss": 0.0006,
      "step": 7720
    },
    {
      "epoch": 0.773,
      "grad_norm": 0.006153833121061325,
      "learning_rate": 1.1350000000000001e-05,
      "loss": 0.0006,
      "step": 7730
    },
    {
      "epoch": 0.774,
      "grad_norm": 0.007491481024771929,
      "learning_rate": 1.13e-05,
      "loss": 0.0007,
      "step": 7740
    },
    {
      "epoch": 0.775,
      "grad_norm": 0.01078925933688879,
      "learning_rate": 1.125e-05,
      "loss": 0.0006,
      "step": 7750
    },
    {
      "epoch": 0.776,
      "grad_norm": 0.005885554943233728,
      "learning_rate": 1.1200000000000001e-05,
      "loss": 0.0006,
      "step": 7760
    },
    {
      "epoch": 0.777,
      "grad_norm": 0.005423078313469887,
      "learning_rate": 1.115e-05,
      "loss": 0.0007,
      "step": 7770
    },
    {
      "epoch": 0.778,
      "grad_norm": 0.008044522255659103,
      "learning_rate": 1.11e-05,
      "loss": 0.0006,
      "step": 7780
    },
    {
      "epoch": 0.779,
      "grad_norm": 0.00733207818120718,
      "learning_rate": 1.1050000000000001e-05,
      "loss": 0.0007,
      "step": 7790
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.0066906120628118515,
      "learning_rate": 1.1000000000000001e-05,
      "loss": 0.0009,
      "step": 7800
    },
    {
      "epoch": 0.781,
      "grad_norm": 0.004443836398422718,
      "learning_rate": 1.095e-05,
      "loss": 0.0006,
      "step": 7810
    },
    {
      "epoch": 0.782,
      "grad_norm": 0.0058379145339131355,
      "learning_rate": 1.09e-05,
      "loss": 0.0007,
      "step": 7820
    },
    {
      "epoch": 0.783,
      "grad_norm": 0.006808693055063486,
      "learning_rate": 1.0850000000000001e-05,
      "loss": 0.0006,
      "step": 7830
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.008773542940616608,
      "learning_rate": 1.08e-05,
      "loss": 0.0006,
      "step": 7840
    },
    {
      "epoch": 0.785,
      "grad_norm": 0.006700740661472082,
      "learning_rate": 1.075e-05,
      "loss": 0.0006,
      "step": 7850
    },
    {
      "epoch": 0.786,
      "grad_norm": 0.00906393863260746,
      "learning_rate": 1.0700000000000001e-05,
      "loss": 0.0006,
      "step": 7860
    },
    {
      "epoch": 0.787,
      "grad_norm": 0.0030822190456092358,
      "learning_rate": 1.065e-05,
      "loss": 0.0005,
      "step": 7870
    },
    {
      "epoch": 0.788,
      "grad_norm": 0.0029632148798555136,
      "learning_rate": 1.06e-05,
      "loss": 0.0005,
      "step": 7880
    },
    {
      "epoch": 0.789,
      "grad_norm": 0.004798842128366232,
      "learning_rate": 1.055e-05,
      "loss": 0.0006,
      "step": 7890
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.007376812864094973,
      "learning_rate": 1.05e-05,
      "loss": 0.0005,
      "step": 7900
    },
    {
      "epoch": 0.791,
      "grad_norm": 0.009337624534964561,
      "learning_rate": 1.045e-05,
      "loss": 0.0009,
      "step": 7910
    },
    {
      "epoch": 0.792,
      "grad_norm": 0.012847904115915298,
      "learning_rate": 1.04e-05,
      "loss": 0.0008,
      "step": 7920
    },
    {
      "epoch": 0.793,
      "grad_norm": 0.005587203428149223,
      "learning_rate": 1.035e-05,
      "loss": 0.0006,
      "step": 7930
    },
    {
      "epoch": 0.794,
      "grad_norm": 0.008464600890874863,
      "learning_rate": 1.03e-05,
      "loss": 0.0006,
      "step": 7940
    },
    {
      "epoch": 0.795,
      "grad_norm": 0.2516852617263794,
      "learning_rate": 1.025e-05,
      "loss": 0.002,
      "step": 7950
    },
    {
      "epoch": 0.796,
      "grad_norm": 0.04664693772792816,
      "learning_rate": 1.02e-05,
      "loss": 0.002,
      "step": 7960
    },
    {
      "epoch": 0.797,
      "grad_norm": 0.02456306852400303,
      "learning_rate": 1.0150000000000001e-05,
      "loss": 0.0013,
      "step": 7970
    },
    {
      "epoch": 0.798,
      "grad_norm": 0.011320951394736767,
      "learning_rate": 1.0100000000000002e-05,
      "loss": 0.0009,
      "step": 7980
    },
    {
      "epoch": 0.799,
      "grad_norm": 0.01860683411359787,
      "learning_rate": 1.005e-05,
      "loss": 0.0012,
      "step": 7990
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.03227970749139786,
      "learning_rate": 1e-05,
      "loss": 0.0009,
      "step": 8000
    },
    {
      "epoch": 0.801,
      "grad_norm": 0.015873363241553307,
      "learning_rate": 9.950000000000001e-06,
      "loss": 0.0008,
      "step": 8010
    },
    {
      "epoch": 0.802,
      "grad_norm": 0.005454899277538061,
      "learning_rate": 9.900000000000002e-06,
      "loss": 0.0008,
      "step": 8020
    },
    {
      "epoch": 0.803,
      "grad_norm": 0.007948348298668861,
      "learning_rate": 9.85e-06,
      "loss": 0.0007,
      "step": 8030
    },
    {
      "epoch": 0.804,
      "grad_norm": 0.013328757137060165,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.0006,
      "step": 8040
    },
    {
      "epoch": 0.805,
      "grad_norm": 0.01018743496388197,
      "learning_rate": 9.750000000000002e-06,
      "loss": 0.0012,
      "step": 8050
    },
    {
      "epoch": 0.806,
      "grad_norm": 0.009421809576451778,
      "learning_rate": 9.7e-06,
      "loss": 0.0008,
      "step": 8060
    },
    {
      "epoch": 0.807,
      "grad_norm": 0.005202045664191246,
      "learning_rate": 9.65e-06,
      "loss": 0.0007,
      "step": 8070
    },
    {
      "epoch": 0.808,
      "grad_norm": 0.012956002727150917,
      "learning_rate": 9.600000000000001e-06,
      "loss": 0.0007,
      "step": 8080
    },
    {
      "epoch": 0.809,
      "grad_norm": 0.006403383333235979,
      "learning_rate": 9.55e-06,
      "loss": 0.0007,
      "step": 8090
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.027560915797948837,
      "learning_rate": 9.5e-06,
      "loss": 0.0008,
      "step": 8100
    },
    {
      "epoch": 0.811,
      "grad_norm": 0.005196988116949797,
      "learning_rate": 9.450000000000001e-06,
      "loss": 0.0006,
      "step": 8110
    },
    {
      "epoch": 0.812,
      "grad_norm": 0.009510821662843227,
      "learning_rate": 9.4e-06,
      "loss": 0.0006,
      "step": 8120
    },
    {
      "epoch": 0.813,
      "grad_norm": 0.006430651992559433,
      "learning_rate": 9.35e-06,
      "loss": 0.0006,
      "step": 8130
    },
    {
      "epoch": 0.814,
      "grad_norm": 0.019426727667450905,
      "learning_rate": 9.3e-06,
      "loss": 0.0009,
      "step": 8140
    },
    {
      "epoch": 0.815,
      "grad_norm": 0.011564865708351135,
      "learning_rate": 9.25e-06,
      "loss": 0.0006,
      "step": 8150
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.009036659263074398,
      "learning_rate": 9.2e-06,
      "loss": 0.0008,
      "step": 8160
    },
    {
      "epoch": 0.817,
      "grad_norm": 0.006685588974505663,
      "learning_rate": 9.15e-06,
      "loss": 0.0007,
      "step": 8170
    },
    {
      "epoch": 0.818,
      "grad_norm": 0.005980687215924263,
      "learning_rate": 9.100000000000001e-06,
      "loss": 0.0005,
      "step": 8180
    },
    {
      "epoch": 0.819,
      "grad_norm": 0.0029402158688753843,
      "learning_rate": 9.05e-06,
      "loss": 0.0005,
      "step": 8190
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.0034720194526016712,
      "learning_rate": 9e-06,
      "loss": 0.0006,
      "step": 8200
    },
    {
      "epoch": 0.821,
      "grad_norm": 0.008967465721070766,
      "learning_rate": 8.95e-06,
      "loss": 0.0009,
      "step": 8210
    },
    {
      "epoch": 0.822,
      "grad_norm": 0.007418784312903881,
      "learning_rate": 8.9e-06,
      "loss": 0.0007,
      "step": 8220
    },
    {
      "epoch": 0.823,
      "grad_norm": 0.0077253603376448154,
      "learning_rate": 8.85e-06,
      "loss": 0.0006,
      "step": 8230
    },
    {
      "epoch": 0.824,
      "grad_norm": 0.011202674359083176,
      "learning_rate": 8.8e-06,
      "loss": 0.0013,
      "step": 8240
    },
    {
      "epoch": 0.825,
      "grad_norm": 0.022354573011398315,
      "learning_rate": 8.75e-06,
      "loss": 0.0008,
      "step": 8250
    },
    {
      "epoch": 0.826,
      "grad_norm": 0.01750505343079567,
      "learning_rate": 8.7e-06,
      "loss": 0.0013,
      "step": 8260
    },
    {
      "epoch": 0.827,
      "grad_norm": 0.01153852604329586,
      "learning_rate": 8.65e-06,
      "loss": 0.0009,
      "step": 8270
    },
    {
      "epoch": 0.828,
      "grad_norm": 0.008752427063882351,
      "learning_rate": 8.599999999999999e-06,
      "loss": 0.0006,
      "step": 8280
    },
    {
      "epoch": 0.829,
      "grad_norm": 0.007307702675461769,
      "learning_rate": 8.550000000000001e-06,
      "loss": 0.0007,
      "step": 8290
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.0077101094648242,
      "learning_rate": 8.500000000000002e-06,
      "loss": 0.0006,
      "step": 8300
    },
    {
      "epoch": 0.831,
      "grad_norm": 0.006358897779136896,
      "learning_rate": 8.45e-06,
      "loss": 0.0005,
      "step": 8310
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.003663134528324008,
      "learning_rate": 8.400000000000001e-06,
      "loss": 0.0006,
      "step": 8320
    },
    {
      "epoch": 0.833,
      "grad_norm": 0.005117372144013643,
      "learning_rate": 8.350000000000001e-06,
      "loss": 0.0006,
      "step": 8330
    },
    {
      "epoch": 0.834,
      "grad_norm": 0.004245636984705925,
      "learning_rate": 8.3e-06,
      "loss": 0.0005,
      "step": 8340
    },
    {
      "epoch": 0.835,
      "grad_norm": 0.005357146263122559,
      "learning_rate": 8.25e-06,
      "loss": 0.0006,
      "step": 8350
    },
    {
      "epoch": 0.836,
      "grad_norm": 0.01055213250219822,
      "learning_rate": 8.200000000000001e-06,
      "loss": 0.0008,
      "step": 8360
    },
    {
      "epoch": 0.837,
      "grad_norm": 0.01871907152235508,
      "learning_rate": 8.15e-06,
      "loss": 0.0007,
      "step": 8370
    },
    {
      "epoch": 0.838,
      "grad_norm": 0.013110162690281868,
      "learning_rate": 8.1e-06,
      "loss": 0.0005,
      "step": 8380
    },
    {
      "epoch": 0.839,
      "grad_norm": 0.005271353758871555,
      "learning_rate": 8.050000000000001e-06,
      "loss": 0.0007,
      "step": 8390
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.004324494861066341,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0005,
      "step": 8400
    },
    {
      "epoch": 0.841,
      "grad_norm": 0.0031851409003138542,
      "learning_rate": 7.95e-06,
      "loss": 0.0006,
      "step": 8410
    },
    {
      "epoch": 0.842,
      "grad_norm": 0.009736557491123676,
      "learning_rate": 7.9e-06,
      "loss": 0.0006,
      "step": 8420
    },
    {
      "epoch": 0.843,
      "grad_norm": 0.005168536212295294,
      "learning_rate": 7.850000000000001e-06,
      "loss": 0.0005,
      "step": 8430
    },
    {
      "epoch": 0.844,
      "grad_norm": 0.002579685300588608,
      "learning_rate": 7.8e-06,
      "loss": 0.0005,
      "step": 8440
    },
    {
      "epoch": 0.845,
      "grad_norm": 0.008710252121090889,
      "learning_rate": 7.75e-06,
      "loss": 0.0005,
      "step": 8450
    },
    {
      "epoch": 0.846,
      "grad_norm": 0.004952189512550831,
      "learning_rate": 7.7e-06,
      "loss": 0.0008,
      "step": 8460
    },
    {
      "epoch": 0.847,
      "grad_norm": 0.003375423140823841,
      "learning_rate": 7.65e-06,
      "loss": 0.0005,
      "step": 8470
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.13184253871440887,
      "learning_rate": 7.6e-06,
      "loss": 0.0012,
      "step": 8480
    },
    {
      "epoch": 0.849,
      "grad_norm": 0.017549166455864906,
      "learning_rate": 7.55e-06,
      "loss": 0.0007,
      "step": 8490
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.00852286908775568,
      "learning_rate": 7.5e-06,
      "loss": 0.0006,
      "step": 8500
    },
    {
      "epoch": 0.851,
      "grad_norm": 0.005547389388084412,
      "learning_rate": 7.45e-06,
      "loss": 0.0005,
      "step": 8510
    },
    {
      "epoch": 0.852,
      "grad_norm": 0.0061622606590390205,
      "learning_rate": 7.4e-06,
      "loss": 0.0005,
      "step": 8520
    },
    {
      "epoch": 0.853,
      "grad_norm": 0.005182339809834957,
      "learning_rate": 7.35e-06,
      "loss": 0.0008,
      "step": 8530
    },
    {
      "epoch": 0.854,
      "grad_norm": 0.005366960074752569,
      "learning_rate": 7.2999999999999996e-06,
      "loss": 0.0006,
      "step": 8540
    },
    {
      "epoch": 0.855,
      "grad_norm": 0.005542315077036619,
      "learning_rate": 7.25e-06,
      "loss": 0.0006,
      "step": 8550
    },
    {
      "epoch": 0.856,
      "grad_norm": 0.003940809518098831,
      "learning_rate": 7.2e-06,
      "loss": 0.0005,
      "step": 8560
    },
    {
      "epoch": 0.857,
      "grad_norm": 0.003730529686436057,
      "learning_rate": 7.15e-06,
      "loss": 0.0006,
      "step": 8570
    },
    {
      "epoch": 0.858,
      "grad_norm": 0.0033961348235607147,
      "learning_rate": 7.1e-06,
      "loss": 0.0005,
      "step": 8580
    },
    {
      "epoch": 0.859,
      "grad_norm": 0.004546662792563438,
      "learning_rate": 7.049999999999999e-06,
      "loss": 0.0006,
      "step": 8590
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.009168008342385292,
      "learning_rate": 7.000000000000001e-06,
      "loss": 0.0005,
      "step": 8600
    },
    {
      "epoch": 0.861,
      "grad_norm": 0.008373426273465157,
      "learning_rate": 6.950000000000001e-06,
      "loss": 0.0008,
      "step": 8610
    },
    {
      "epoch": 0.862,
      "grad_norm": 0.004947313107550144,
      "learning_rate": 6.900000000000001e-06,
      "loss": 0.0006,
      "step": 8620
    },
    {
      "epoch": 0.863,
      "grad_norm": 0.015127859078347683,
      "learning_rate": 6.8500000000000005e-06,
      "loss": 0.0006,
      "step": 8630
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.0056435600854456425,
      "learning_rate": 6.800000000000001e-06,
      "loss": 0.0006,
      "step": 8640
    },
    {
      "epoch": 0.865,
      "grad_norm": 0.004109732341021299,
      "learning_rate": 6.750000000000001e-06,
      "loss": 0.0005,
      "step": 8650
    },
    {
      "epoch": 0.866,
      "grad_norm": 0.006170314736664295,
      "learning_rate": 6.700000000000001e-06,
      "loss": 0.0005,
      "step": 8660
    },
    {
      "epoch": 0.867,
      "grad_norm": 0.002802550094202161,
      "learning_rate": 6.650000000000001e-06,
      "loss": 0.0005,
      "step": 8670
    },
    {
      "epoch": 0.868,
      "grad_norm": 0.0029788350220769644,
      "learning_rate": 6.6e-06,
      "loss": 0.0004,
      "step": 8680
    },
    {
      "epoch": 0.869,
      "grad_norm": 0.013022363185882568,
      "learning_rate": 6.550000000000001e-06,
      "loss": 0.0006,
      "step": 8690
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.0036853367928415537,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 0.0006,
      "step": 8700
    },
    {
      "epoch": 0.871,
      "grad_norm": 0.002578242914751172,
      "learning_rate": 6.45e-06,
      "loss": 0.0005,
      "step": 8710
    },
    {
      "epoch": 0.872,
      "grad_norm": 0.0036895396187901497,
      "learning_rate": 6.4000000000000006e-06,
      "loss": 0.0005,
      "step": 8720
    },
    {
      "epoch": 0.873,
      "grad_norm": 0.006020987406373024,
      "learning_rate": 6.35e-06,
      "loss": 0.0005,
      "step": 8730
    },
    {
      "epoch": 0.874,
      "grad_norm": 0.006671608425676823,
      "learning_rate": 6.300000000000001e-06,
      "loss": 0.0006,
      "step": 8740
    },
    {
      "epoch": 0.875,
      "grad_norm": 0.0038102639373391867,
      "learning_rate": 6.25e-06,
      "loss": 0.0006,
      "step": 8750
    },
    {
      "epoch": 0.876,
      "grad_norm": 0.006786294747143984,
      "learning_rate": 6.2e-06,
      "loss": 0.0004,
      "step": 8760
    },
    {
      "epoch": 0.877,
      "grad_norm": 0.00381205091252923,
      "learning_rate": 6.15e-06,
      "loss": 0.0004,
      "step": 8770
    },
    {
      "epoch": 0.878,
      "grad_norm": 0.007368630729615688,
      "learning_rate": 6.1e-06,
      "loss": 0.0005,
      "step": 8780
    },
    {
      "epoch": 0.879,
      "grad_norm": 0.0035172586794942617,
      "learning_rate": 6.0500000000000005e-06,
      "loss": 0.0006,
      "step": 8790
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.005555720068514347,
      "learning_rate": 6e-06,
      "loss": 0.0007,
      "step": 8800
    },
    {
      "epoch": 0.881,
      "grad_norm": 0.0076825893484056,
      "learning_rate": 5.95e-06,
      "loss": 0.0005,
      "step": 8810
    },
    {
      "epoch": 0.882,
      "grad_norm": 0.0055446140468120575,
      "learning_rate": 5.9e-06,
      "loss": 0.0005,
      "step": 8820
    },
    {
      "epoch": 0.883,
      "grad_norm": 0.002265618182718754,
      "learning_rate": 5.850000000000001e-06,
      "loss": 0.0005,
      "step": 8830
    },
    {
      "epoch": 0.884,
      "grad_norm": 0.003428585361689329,
      "learning_rate": 5.8e-06,
      "loss": 0.0004,
      "step": 8840
    },
    {
      "epoch": 0.885,
      "grad_norm": 0.0044764927588403225,
      "learning_rate": 5.750000000000001e-06,
      "loss": 0.0005,
      "step": 8850
    },
    {
      "epoch": 0.886,
      "grad_norm": 0.003201392712071538,
      "learning_rate": 5.7000000000000005e-06,
      "loss": 0.0005,
      "step": 8860
    },
    {
      "epoch": 0.887,
      "grad_norm": 0.0029762780759483576,
      "learning_rate": 5.65e-06,
      "loss": 0.0006,
      "step": 8870
    },
    {
      "epoch": 0.888,
      "grad_norm": 0.07450267672538757,
      "learning_rate": 5.600000000000001e-06,
      "loss": 0.0009,
      "step": 8880
    },
    {
      "epoch": 0.889,
      "grad_norm": 0.006392148323357105,
      "learning_rate": 5.55e-06,
      "loss": 0.0006,
      "step": 8890
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.0038995451759546995,
      "learning_rate": 5.500000000000001e-06,
      "loss": 0.0005,
      "step": 8900
    },
    {
      "epoch": 0.891,
      "grad_norm": 0.0028438065201044083,
      "learning_rate": 5.45e-06,
      "loss": 0.0004,
      "step": 8910
    },
    {
      "epoch": 0.892,
      "grad_norm": 0.003168331226333976,
      "learning_rate": 5.4e-06,
      "loss": 0.0004,
      "step": 8920
    },
    {
      "epoch": 0.893,
      "grad_norm": 0.0026163198053836823,
      "learning_rate": 5.3500000000000004e-06,
      "loss": 0.0004,
      "step": 8930
    },
    {
      "epoch": 0.894,
      "grad_norm": 0.0029086521826684475,
      "learning_rate": 5.3e-06,
      "loss": 0.0005,
      "step": 8940
    },
    {
      "epoch": 0.895,
      "grad_norm": 0.011433840729296207,
      "learning_rate": 5.25e-06,
      "loss": 0.0007,
      "step": 8950
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.01782575435936451,
      "learning_rate": 5.2e-06,
      "loss": 0.0011,
      "step": 8960
    },
    {
      "epoch": 0.897,
      "grad_norm": 0.00613692682236433,
      "learning_rate": 5.15e-06,
      "loss": 0.0004,
      "step": 8970
    },
    {
      "epoch": 0.898,
      "grad_norm": 0.02408697083592415,
      "learning_rate": 5.1e-06,
      "loss": 0.0007,
      "step": 8980
    },
    {
      "epoch": 0.899,
      "grad_norm": 0.004028539173305035,
      "learning_rate": 5.050000000000001e-06,
      "loss": 0.0005,
      "step": 8990
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.0032080088276416063,
      "learning_rate": 5e-06,
      "loss": 0.0005,
      "step": 9000
    },
    {
      "epoch": 0.901,
      "grad_norm": 0.0035681568551808596,
      "learning_rate": 4.950000000000001e-06,
      "loss": 0.0004,
      "step": 9010
    },
    {
      "epoch": 0.902,
      "grad_norm": 0.007591512985527515,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0005,
      "step": 9020
    },
    {
      "epoch": 0.903,
      "grad_norm": 0.004855870269238949,
      "learning_rate": 4.85e-06,
      "loss": 0.0004,
      "step": 9030
    },
    {
      "epoch": 0.904,
      "grad_norm": 0.004854188766330481,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0004,
      "step": 9040
    },
    {
      "epoch": 0.905,
      "grad_norm": 0.004117886070162058,
      "learning_rate": 4.75e-06,
      "loss": 0.0005,
      "step": 9050
    },
    {
      "epoch": 0.906,
      "grad_norm": 0.0045243133790791035,
      "learning_rate": 4.7e-06,
      "loss": 0.0005,
      "step": 9060
    },
    {
      "epoch": 0.907,
      "grad_norm": 0.001863984507508576,
      "learning_rate": 4.65e-06,
      "loss": 0.0004,
      "step": 9070
    },
    {
      "epoch": 0.908,
      "grad_norm": 0.002472365740686655,
      "learning_rate": 4.6e-06,
      "loss": 0.0005,
      "step": 9080
    },
    {
      "epoch": 0.909,
      "grad_norm": 0.0020466954447329044,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0004,
      "step": 9090
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.004180034622550011,
      "learning_rate": 4.5e-06,
      "loss": 0.0004,
      "step": 9100
    },
    {
      "epoch": 0.911,
      "grad_norm": 0.00341266137547791,
      "learning_rate": 4.45e-06,
      "loss": 0.0006,
      "step": 9110
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.006567875389009714,
      "learning_rate": 4.4e-06,
      "loss": 0.0004,
      "step": 9120
    },
    {
      "epoch": 0.913,
      "grad_norm": 0.003975498490035534,
      "learning_rate": 4.35e-06,
      "loss": 0.0006,
      "step": 9130
    },
    {
      "epoch": 0.914,
      "grad_norm": 0.003391894046217203,
      "learning_rate": 4.2999999999999995e-06,
      "loss": 0.0006,
      "step": 9140
    },
    {
      "epoch": 0.915,
      "grad_norm": 0.005821021273732185,
      "learning_rate": 4.250000000000001e-06,
      "loss": 0.0004,
      "step": 9150
    },
    {
      "epoch": 0.916,
      "grad_norm": 0.0022448371164500713,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0004,
      "step": 9160
    },
    {
      "epoch": 0.917,
      "grad_norm": 0.003718709573149681,
      "learning_rate": 4.15e-06,
      "loss": 0.0004,
      "step": 9170
    },
    {
      "epoch": 0.918,
      "grad_norm": 0.008243223652243614,
      "learning_rate": 4.1000000000000006e-06,
      "loss": 0.0007,
      "step": 9180
    },
    {
      "epoch": 0.919,
      "grad_norm": 0.010773789137601852,
      "learning_rate": 4.05e-06,
      "loss": 0.0007,
      "step": 9190
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.006589268799871206,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0005,
      "step": 9200
    },
    {
      "epoch": 0.921,
      "grad_norm": 0.0026856744661927223,
      "learning_rate": 3.95e-06,
      "loss": 0.0004,
      "step": 9210
    },
    {
      "epoch": 0.922,
      "grad_norm": 0.012134186923503876,
      "learning_rate": 3.9e-06,
      "loss": 0.0005,
      "step": 9220
    },
    {
      "epoch": 0.923,
      "grad_norm": 0.004260225687175989,
      "learning_rate": 3.85e-06,
      "loss": 0.0005,
      "step": 9230
    },
    {
      "epoch": 0.924,
      "grad_norm": 0.0023803950753062963,
      "learning_rate": 3.8e-06,
      "loss": 0.0004,
      "step": 9240
    },
    {
      "epoch": 0.925,
      "grad_norm": 0.0037502460181713104,
      "learning_rate": 3.75e-06,
      "loss": 0.0005,
      "step": 9250
    },
    {
      "epoch": 0.926,
      "grad_norm": 0.0017525887815281749,
      "learning_rate": 3.7e-06,
      "loss": 0.0003,
      "step": 9260
    },
    {
      "epoch": 0.927,
      "grad_norm": 0.003996537532657385,
      "learning_rate": 3.6499999999999998e-06,
      "loss": 0.0005,
      "step": 9270
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.009158821776509285,
      "learning_rate": 3.6e-06,
      "loss": 0.0007,
      "step": 9280
    },
    {
      "epoch": 0.929,
      "grad_norm": 0.003372638253495097,
      "learning_rate": 3.55e-06,
      "loss": 0.0004,
      "step": 9290
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.0026602360885590315,
      "learning_rate": 3.5000000000000004e-06,
      "loss": 0.0004,
      "step": 9300
    },
    {
      "epoch": 0.931,
      "grad_norm": 0.014532738365232944,
      "learning_rate": 3.4500000000000004e-06,
      "loss": 0.0007,
      "step": 9310
    },
    {
      "epoch": 0.932,
      "grad_norm": 0.002912462456151843,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0004,
      "step": 9320
    },
    {
      "epoch": 0.933,
      "grad_norm": 0.0052029709331691265,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0006,
      "step": 9330
    },
    {
      "epoch": 0.934,
      "grad_norm": 0.016220854595303535,
      "learning_rate": 3.3e-06,
      "loss": 0.0004,
      "step": 9340
    },
    {
      "epoch": 0.935,
      "grad_norm": 0.0030162036418914795,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0004,
      "step": 9350
    },
    {
      "epoch": 0.936,
      "grad_norm": 0.002491691382601857,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.0004,
      "step": 9360
    },
    {
      "epoch": 0.937,
      "grad_norm": 0.022630969062447548,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0005,
      "step": 9370
    },
    {
      "epoch": 0.938,
      "grad_norm": 0.005951160565018654,
      "learning_rate": 3.1e-06,
      "loss": 0.0005,
      "step": 9380
    },
    {
      "epoch": 0.939,
      "grad_norm": 0.0024763622786849737,
      "learning_rate": 3.05e-06,
      "loss": 0.0005,
      "step": 9390
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.0049979290924966335,
      "learning_rate": 3e-06,
      "loss": 0.0005,
      "step": 9400
    },
    {
      "epoch": 0.941,
      "grad_norm": 0.0025999436620622873,
      "learning_rate": 2.95e-06,
      "loss": 0.0004,
      "step": 9410
    },
    {
      "epoch": 0.942,
      "grad_norm": 0.004584169946610928,
      "learning_rate": 2.9e-06,
      "loss": 0.0006,
      "step": 9420
    },
    {
      "epoch": 0.943,
      "grad_norm": 0.005211680196225643,
      "learning_rate": 2.8500000000000002e-06,
      "loss": 0.0005,
      "step": 9430
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.0022507943212985992,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0004,
      "step": 9440
    },
    {
      "epoch": 0.945,
      "grad_norm": 0.0029024691320955753,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0004,
      "step": 9450
    },
    {
      "epoch": 0.946,
      "grad_norm": 0.003968573175370693,
      "learning_rate": 2.7e-06,
      "loss": 0.0004,
      "step": 9460
    },
    {
      "epoch": 0.947,
      "grad_norm": 0.003264777595177293,
      "learning_rate": 2.65e-06,
      "loss": 0.0005,
      "step": 9470
    },
    {
      "epoch": 0.948,
      "grad_norm": 0.0048127188347280025,
      "learning_rate": 2.6e-06,
      "loss": 0.0004,
      "step": 9480
    },
    {
      "epoch": 0.949,
      "grad_norm": 0.004405410494655371,
      "learning_rate": 2.55e-06,
      "loss": 0.0006,
      "step": 9490
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.00462340796366334,
      "learning_rate": 2.5e-06,
      "loss": 0.0004,
      "step": 9500
    },
    {
      "epoch": 0.951,
      "grad_norm": 0.0021721452940255404,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0004,
      "step": 9510
    },
    {
      "epoch": 0.952,
      "grad_norm": 0.002355078933760524,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0006,
      "step": 9520
    },
    {
      "epoch": 0.953,
      "grad_norm": 0.0022414589766412973,
      "learning_rate": 2.35e-06,
      "loss": 0.0004,
      "step": 9530
    },
    {
      "epoch": 0.954,
      "grad_norm": 0.012005253694951534,
      "learning_rate": 2.3e-06,
      "loss": 0.0006,
      "step": 9540
    },
    {
      "epoch": 0.955,
      "grad_norm": 0.00513832364231348,
      "learning_rate": 2.25e-06,
      "loss": 0.0005,
      "step": 9550
    },
    {
      "epoch": 0.956,
      "grad_norm": 0.0027625642251223326,
      "learning_rate": 2.2e-06,
      "loss": 0.0005,
      "step": 9560
    },
    {
      "epoch": 0.957,
      "grad_norm": 0.008645957335829735,
      "learning_rate": 2.1499999999999997e-06,
      "loss": 0.0005,
      "step": 9570
    },
    {
      "epoch": 0.958,
      "grad_norm": 0.00188863230869174,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.0004,
      "step": 9580
    },
    {
      "epoch": 0.959,
      "grad_norm": 0.0025561931543052197,
      "learning_rate": 2.0500000000000003e-06,
      "loss": 0.0004,
      "step": 9590
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.0033618698362261057,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0004,
      "step": 9600
    },
    {
      "epoch": 0.961,
      "grad_norm": 0.0018735548947006464,
      "learning_rate": 1.95e-06,
      "loss": 0.0004,
      "step": 9610
    },
    {
      "epoch": 0.962,
      "grad_norm": 0.0019359014695510268,
      "learning_rate": 1.9e-06,
      "loss": 0.0005,
      "step": 9620
    },
    {
      "epoch": 0.963,
      "grad_norm": 0.005369434133172035,
      "learning_rate": 1.85e-06,
      "loss": 0.0004,
      "step": 9630
    },
    {
      "epoch": 0.964,
      "grad_norm": 0.0017576682148501277,
      "learning_rate": 1.8e-06,
      "loss": 0.0004,
      "step": 9640
    },
    {
      "epoch": 0.965,
      "grad_norm": 0.002633103635162115,
      "learning_rate": 1.7500000000000002e-06,
      "loss": 0.0004,
      "step": 9650
    },
    {
      "epoch": 0.966,
      "grad_norm": 0.007023205049335957,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.0004,
      "step": 9660
    },
    {
      "epoch": 0.967,
      "grad_norm": 0.0026062438264489174,
      "learning_rate": 1.65e-06,
      "loss": 0.0005,
      "step": 9670
    },
    {
      "epoch": 0.968,
      "grad_norm": 0.0025111304130405188,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.0005,
      "step": 9680
    },
    {
      "epoch": 0.969,
      "grad_norm": 0.0028218806255608797,
      "learning_rate": 1.55e-06,
      "loss": 0.0004,
      "step": 9690
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.0024802633561193943,
      "learning_rate": 1.5e-06,
      "loss": 0.0005,
      "step": 9700
    },
    {
      "epoch": 0.971,
      "grad_norm": 0.002883223118260503,
      "learning_rate": 1.45e-06,
      "loss": 0.0005,
      "step": 9710
    },
    {
      "epoch": 0.972,
      "grad_norm": 0.002503247233107686,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.0005,
      "step": 9720
    },
    {
      "epoch": 0.973,
      "grad_norm": 0.002495008986443281,
      "learning_rate": 1.35e-06,
      "loss": 0.0006,
      "step": 9730
    },
    {
      "epoch": 0.974,
      "grad_norm": 0.03429775312542915,
      "learning_rate": 1.3e-06,
      "loss": 0.0006,
      "step": 9740
    },
    {
      "epoch": 0.975,
      "grad_norm": 0.003482217201963067,
      "learning_rate": 1.25e-06,
      "loss": 0.0004,
      "step": 9750
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.001837963704019785,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 0.0004,
      "step": 9760
    },
    {
      "epoch": 0.977,
      "grad_norm": 0.0020507893059402704,
      "learning_rate": 1.15e-06,
      "loss": 0.0004,
      "step": 9770
    },
    {
      "epoch": 0.978,
      "grad_norm": 0.0022647210862487555,
      "learning_rate": 1.1e-06,
      "loss": 0.0005,
      "step": 9780
    },
    {
      "epoch": 0.979,
      "grad_norm": 0.0017425378318876028,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 0.0004,
      "step": 9790
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2319187968969345,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 0.0021,
      "step": 9800
    },
    {
      "epoch": 0.981,
      "grad_norm": 0.01799739897251129,
      "learning_rate": 9.5e-07,
      "loss": 0.0006,
      "step": 9810
    },
    {
      "epoch": 0.982,
      "grad_norm": 0.007147952448576689,
      "learning_rate": 9e-07,
      "loss": 0.0005,
      "step": 9820
    },
    {
      "epoch": 0.983,
      "grad_norm": 0.004181794356554747,
      "learning_rate": 8.500000000000001e-07,
      "loss": 0.0005,
      "step": 9830
    },
    {
      "epoch": 0.984,
      "grad_norm": 0.00277232495136559,
      "learning_rate": 8.000000000000001e-07,
      "loss": 0.0004,
      "step": 9840
    },
    {
      "epoch": 0.985,
      "grad_norm": 0.0024797001387923956,
      "learning_rate": 7.5e-07,
      "loss": 0.0006,
      "step": 9850
    },
    {
      "epoch": 0.986,
      "grad_norm": 0.002748242113739252,
      "learning_rate": 7.000000000000001e-07,
      "loss": 0.0005,
      "step": 9860
    },
    {
      "epoch": 0.987,
      "grad_norm": 0.002988820429891348,
      "learning_rate": 6.5e-07,
      "loss": 0.0004,
      "step": 9870
    },
    {
      "epoch": 0.988,
      "grad_norm": 0.002272873418405652,
      "learning_rate": 6.000000000000001e-07,
      "loss": 0.0006,
      "step": 9880
    },
    {
      "epoch": 0.989,
      "grad_norm": 0.0028824047185480595,
      "learning_rate": 5.5e-07,
      "loss": 0.0005,
      "step": 9890
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.013895529322326183,
      "learning_rate": 5.000000000000001e-07,
      "loss": 0.0005,
      "step": 9900
    },
    {
      "epoch": 0.991,
      "grad_norm": 0.004210934974253178,
      "learning_rate": 4.5e-07,
      "loss": 0.0004,
      "step": 9910
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.0017349456902593374,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 0.0005,
      "step": 9920
    },
    {
      "epoch": 0.993,
      "grad_norm": 0.0036622195038944483,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 0.0003,
      "step": 9930
    },
    {
      "epoch": 0.994,
      "grad_norm": 0.02928483486175537,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 0.0006,
      "step": 9940
    },
    {
      "epoch": 0.995,
      "grad_norm": 0.004271595273166895,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 0.0004,
      "step": 9950
    },
    {
      "epoch": 0.996,
      "grad_norm": 0.004935207776725292,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 0.0004,
      "step": 9960
    },
    {
      "epoch": 0.997,
      "grad_norm": 0.005258087068796158,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 0.0004,
      "step": 9970
    },
    {
      "epoch": 0.998,
      "grad_norm": 0.0014150363858789206,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 0.0004,
      "step": 9980
    },
    {
      "epoch": 0.999,
      "grad_norm": 0.003183445893228054,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 0.0004,
      "step": 9990
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0063432566821575165,
      "learning_rate": 0.0,
      "loss": 0.0004,
      "step": 10000
    }
  ],
  "logging_steps": 10,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.6962203336704e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}