{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.666250260254008,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "acc": 0.0,
      "epoch": 0.0,
      "learning_rate": 5e-06,
      "loss": 1.0061,
      "rl_loss": 0.0004,
      "step": 1
    },
    {
      "acc": 0.0,
      "epoch": 0.01,
      "learning_rate": 5e-06,
      "loss": 0.3002,
      "rl_loss": 0.0001,
      "step": 10
    },
    {
      "acc": 0.0,
      "epoch": 0.01,
      "learning_rate": 5e-06,
      "loss": 0.2233,
      "rl_loss": 0.0004,
      "step": 20
    },
    {
      "acc": 0.0,
      "epoch": 0.02,
      "learning_rate": 5e-06,
      "loss": 0.2344,
      "rl_loss": 0.0,
      "step": 30
    },
    {
      "acc": 0.0,
      "epoch": 0.03,
      "learning_rate": 5e-06,
      "loss": 0.2459,
      "rl_loss": 0.0001,
      "step": 40
    },
    {
      "acc": 0.0,
      "epoch": 0.03,
      "learning_rate": 5e-06,
      "loss": 0.2099,
      "rl_loss": 0.0002,
      "step": 50
    },
    {
      "acc": 0.0,
      "epoch": 0.04,
      "learning_rate": 5e-06,
      "loss": 0.2116,
      "rl_loss": 0.0,
      "step": 60
    },
    {
      "acc": 0.0,
      "epoch": 0.05,
      "learning_rate": 5e-06,
      "loss": 0.2028,
      "rl_loss": 0.0001,
      "step": 70
    },
    {
      "acc": 0.0,
      "epoch": 0.05,
      "learning_rate": 5e-06,
      "loss": 0.2669,
      "rl_loss": 0.0001,
      "step": 80
    },
    {
      "acc": 0.0,
      "epoch": 0.06,
      "learning_rate": 5e-06,
      "loss": 0.2903,
      "rl_loss": 0.0002,
      "step": 90
    },
    {
      "acc": 0.0,
      "epoch": 0.07,
      "learning_rate": 5e-06,
      "loss": 0.2818,
      "rl_loss": 0.0001,
      "step": 100
    },
    {
      "acc": 0.0,
      "epoch": 0.07,
      "learning_rate": 5e-06,
      "loss": 0.2464,
      "rl_loss": 0.0003,
      "step": 110
    },
    {
      "acc": 0.0,
      "epoch": 0.08,
      "learning_rate": 5e-06,
      "loss": 0.2283,
      "rl_loss": 0.0002,
      "step": 120
    },
    {
      "acc": 0.0,
      "epoch": 0.09,
      "learning_rate": 5e-06,
      "loss": 0.2443,
      "rl_loss": 0.0002,
      "step": 130
    },
    {
      "acc": 0.0,
      "epoch": 0.09,
      "learning_rate": 5e-06,
      "loss": 0.231,
      "rl_loss": 0.0004,
      "step": 140
    },
    {
      "acc": 0.0,
      "epoch": 0.1,
      "learning_rate": 5e-06,
      "loss": 0.2179,
      "rl_loss": 0.0002,
      "step": 150
    },
    {
      "acc": 0.0,
      "epoch": 0.11,
      "learning_rate": 5e-06,
      "loss": 0.2054,
      "rl_loss": 0.0,
      "step": 160
    },
    {
      "acc": 0.0,
      "epoch": 0.11,
      "learning_rate": 5e-06,
      "loss": 0.2055,
      "rl_loss": 0.0004,
      "step": 170
    },
    {
      "acc": 0.0,
      "epoch": 0.12,
      "learning_rate": 5e-06,
      "loss": 0.2095,
      "rl_loss": 0.0003,
      "step": 180
    },
    {
      "acc": 0.0,
      "epoch": 0.13,
      "learning_rate": 5e-06,
      "loss": 0.2825,
      "rl_loss": 0.0005,
      "step": 190
    },
    {
      "acc": 0.0,
      "epoch": 0.13,
      "learning_rate": 5e-06,
      "loss": 0.1994,
      "rl_loss": 0.0001,
      "step": 200
    },
    {
      "acc": 0.0,
      "epoch": 0.14,
      "learning_rate": 5e-06,
      "loss": 0.2121,
      "rl_loss": 0.0001,
      "step": 210
    },
    {
      "acc": 0.0,
      "epoch": 0.15,
      "learning_rate": 5e-06,
      "loss": 0.2321,
      "rl_loss": 0.0001,
      "step": 220
    },
    {
      "acc": 0.0,
      "epoch": 0.15,
      "learning_rate": 5e-06,
      "loss": 0.2396,
      "rl_loss": 0.0005,
      "step": 230
    },
    {
      "acc": 0.0,
      "epoch": 0.16,
      "learning_rate": 5e-06,
      "loss": 0.209,
      "rl_loss": 0.0001,
      "step": 240
    },
    {
      "acc": 0.0,
      "epoch": 0.17,
      "learning_rate": 5e-06,
      "loss": 0.2048,
      "rl_loss": 0.0001,
      "step": 250
    },
    {
      "acc": 0.0,
      "epoch": 0.17,
      "learning_rate": 5e-06,
      "loss": 0.2018,
      "rl_loss": 0.0001,
      "step": 260
    },
    {
      "acc": 0.0,
      "epoch": 0.18,
      "learning_rate": 5e-06,
      "loss": 0.2088,
      "rl_loss": 0.0002,
      "step": 270
    },
    {
      "acc": 0.0,
      "epoch": 0.19,
      "learning_rate": 5e-06,
      "loss": 0.2048,
      "rl_loss": 0.0002,
      "step": 280
    },
    {
      "acc": 0.0,
      "epoch": 0.19,
      "learning_rate": 5e-06,
      "loss": 0.2075,
      "rl_loss": 0.0001,
      "step": 290
    },
    {
      "acc": 0.0,
      "epoch": 0.2,
      "learning_rate": 5e-06,
      "loss": 0.2085,
      "rl_loss": 0.0003,
      "step": 300
    },
    {
      "acc": 0.0,
      "epoch": 0.21,
      "learning_rate": 5e-06,
      "loss": 0.1934,
      "rl_loss": 0.0001,
      "step": 310
    },
    {
      "acc": 0.0,
      "epoch": 0.21,
      "learning_rate": 5e-06,
      "loss": 0.203,
      "rl_loss": 0.0001,
      "step": 320
    },
    {
      "acc": 0.0,
      "epoch": 0.22,
      "learning_rate": 5e-06,
      "loss": 0.2069,
      "rl_loss": 0.0001,
      "step": 330
    },
    {
      "acc": 0.0,
      "epoch": 0.23,
      "learning_rate": 5e-06,
      "loss": 0.2426,
      "rl_loss": 0.0001,
      "step": 340
    },
    {
      "acc": 0.0,
      "epoch": 0.23,
      "learning_rate": 5e-06,
      "loss": 0.202,
      "rl_loss": 0.0001,
      "step": 350
    },
    {
      "acc": 0.0,
      "epoch": 0.24,
      "learning_rate": 5e-06,
      "loss": 0.2008,
      "rl_loss": 0.0001,
      "step": 360
    },
    {
      "acc": 0.0,
      "epoch": 0.25,
      "learning_rate": 5e-06,
      "loss": 0.1891,
      "rl_loss": 0.0002,
      "step": 370
    },
    {
      "acc": 0.0,
      "epoch": 0.25,
      "learning_rate": 5e-06,
      "loss": 0.1917,
      "rl_loss": 0.0001,
      "step": 380
    },
    {
      "acc": 0.0,
      "epoch": 0.26,
      "learning_rate": 5e-06,
      "loss": 0.2043,
      "rl_loss": 0.0004,
      "step": 390
    },
    {
      "acc": 0.0,
      "epoch": 0.27,
      "learning_rate": 5e-06,
      "loss": 0.2111,
      "rl_loss": 0.0001,
      "step": 400
    },
    {
      "acc": 0.0,
      "epoch": 0.27,
      "learning_rate": 5e-06,
      "loss": 0.1959,
      "rl_loss": 0.0002,
      "step": 410
    },
    {
      "acc": 0.0,
      "epoch": 0.28,
      "learning_rate": 5e-06,
      "loss": 0.2019,
      "rl_loss": 0.0003,
      "step": 420
    },
    {
      "acc": 0.0,
      "epoch": 0.29,
      "learning_rate": 5e-06,
      "loss": 0.2019,
      "rl_loss": 0.0002,
      "step": 430
    },
    {
      "acc": 0.0,
      "epoch": 0.29,
      "learning_rate": 5e-06,
      "loss": 0.2222,
      "rl_loss": 0.0005,
      "step": 440
    },
    {
      "acc": 0.0,
      "epoch": 0.3,
      "learning_rate": 5e-06,
      "loss": 0.2047,
      "rl_loss": 0.0001,
      "step": 450
    },
    {
      "acc": 0.0,
      "epoch": 0.31,
      "learning_rate": 5e-06,
      "loss": 0.2091,
      "rl_loss": 0.0001,
      "step": 460
    },
    {
      "acc": 0.0,
      "epoch": 0.31,
      "learning_rate": 5e-06,
      "loss": 0.1873,
      "rl_loss": 0.0,
      "step": 470
    },
    {
      "acc": 0.0,
      "epoch": 0.32,
      "learning_rate": 5e-06,
      "loss": 0.2177,
      "rl_loss": 0.0001,
      "step": 480
    },
    {
      "acc": 0.0,
      "epoch": 0.33,
      "learning_rate": 5e-06,
      "loss": 0.1975,
      "rl_loss": 0.0004,
      "step": 490
    },
    {
      "acc": 0.0,
      "epoch": 0.33,
      "learning_rate": 5e-06,
      "loss": 0.2283,
      "rl_loss": 0.0002,
      "step": 500
    },
    {
      "acc": 0.0,
      "epoch": 0.34,
      "learning_rate": 5e-06,
      "loss": 0.2169,
      "rl_loss": 0.0,
      "step": 510
    },
    {
      "acc": 0.0,
      "epoch": 0.35,
      "learning_rate": 5e-06,
      "loss": 0.1988,
      "rl_loss": 0.0003,
      "step": 520
    },
    {
      "acc": 0.0,
      "epoch": 0.35,
      "learning_rate": 5e-06,
      "loss": 0.1905,
      "rl_loss": 0.0001,
      "step": 530
    },
    {
      "acc": 0.0,
      "epoch": 0.36,
      "learning_rate": 5e-06,
      "loss": 0.2188,
      "rl_loss": 0.0,
      "step": 540
    },
    {
      "acc": 0.0,
      "epoch": 0.37,
      "learning_rate": 5e-06,
      "loss": 0.2556,
      "rl_loss": 0.0001,
      "step": 550
    },
    {
      "acc": 0.0,
      "epoch": 0.37,
      "learning_rate": 5e-06,
      "loss": 0.217,
      "rl_loss": 0.0,
      "step": 560
    },
    {
      "acc": 0.0,
      "epoch": 0.38,
      "learning_rate": 5e-06,
      "loss": 0.2118,
      "rl_loss": 0.0002,
      "step": 570
    },
    {
      "acc": 0.0,
      "epoch": 0.39,
      "learning_rate": 5e-06,
      "loss": 0.193,
      "rl_loss": 0.0003,
      "step": 580
    },
    {
      "acc": 0.0,
      "epoch": 0.39,
      "learning_rate": 5e-06,
      "loss": 0.2136,
      "rl_loss": 0.0002,
      "step": 590
    },
    {
      "acc": 0.0,
      "epoch": 0.4,
      "learning_rate": 5e-06,
      "loss": 0.2021,
      "rl_loss": 0.0005,
      "step": 600
    },
    {
      "acc": 0.0,
      "epoch": 0.41,
      "learning_rate": 5e-06,
      "loss": 0.1884,
      "rl_loss": 0.0004,
      "step": 610
    },
    {
      "acc": 0.0,
      "epoch": 0.41,
      "learning_rate": 5e-06,
      "loss": 0.1951,
      "rl_loss": 0.0002,
      "step": 620
    },
    {
      "acc": 0.0,
      "epoch": 0.42,
      "learning_rate": 5e-06,
      "loss": 0.1952,
      "rl_loss": 0.0002,
      "step": 630
    },
    {
      "acc": 0.0,
      "epoch": 0.43,
      "learning_rate": 5e-06,
      "loss": 0.2556,
      "rl_loss": 0.0003,
      "step": 640
    },
    {
      "acc": 0.0,
      "epoch": 0.43,
      "learning_rate": 5e-06,
      "loss": 0.1936,
      "rl_loss": 0.0001,
      "step": 650
    },
    {
      "acc": 0.0,
      "epoch": 0.44,
      "learning_rate": 5e-06,
      "loss": 0.2081,
      "rl_loss": 0.0001,
      "step": 660
    },
    {
      "acc": 0.0,
      "epoch": 0.45,
      "learning_rate": 5e-06,
      "loss": 0.1917,
      "rl_loss": 0.0001,
      "step": 670
    },
    {
      "acc": 0.0,
      "epoch": 0.45,
      "learning_rate": 5e-06,
      "loss": 0.198,
      "rl_loss": 0.0001,
      "step": 680
    },
    {
      "acc": 0.0,
      "epoch": 0.46,
      "learning_rate": 5e-06,
      "loss": 0.1886,
      "rl_loss": 0.0003,
      "step": 690
    },
    {
      "acc": 0.0,
      "epoch": 0.47,
      "learning_rate": 5e-06,
      "loss": 0.1974,
      "rl_loss": 0.0003,
      "step": 700
    },
    {
      "acc": 0.0,
      "epoch": 0.47,
      "learning_rate": 5e-06,
      "loss": 0.2088,
      "rl_loss": 0.0004,
      "step": 710
    },
    {
      "acc": 0.0,
      "epoch": 0.48,
      "learning_rate": 5e-06,
      "loss": 0.2018,
      "rl_loss": 0.0001,
      "step": 720
    },
    {
      "acc": 0.0,
      "epoch": 0.49,
      "learning_rate": 5e-06,
      "loss": 0.2074,
      "rl_loss": 0.0,
      "step": 730
    },
    {
      "acc": 0.0,
      "epoch": 0.49,
      "learning_rate": 5e-06,
      "loss": 0.1952,
      "rl_loss": 0.0001,
      "step": 740
    },
    {
      "acc": 0.0,
      "epoch": 0.5,
      "learning_rate": 5e-06,
      "loss": 0.2048,
      "rl_loss": 0.0002,
      "step": 750
    },
    {
      "acc": 0.0,
      "epoch": 0.51,
      "learning_rate": 5e-06,
      "loss": 0.2158,
      "rl_loss": 0.0004,
      "step": 760
    },
    {
      "acc": 0.0,
      "epoch": 0.51,
      "learning_rate": 5e-06,
      "loss": 0.1941,
      "rl_loss": 0.0003,
      "step": 770
    },
    {
      "acc": 0.0,
      "epoch": 0.52,
      "learning_rate": 5e-06,
      "loss": 0.2159,
      "rl_loss": 0.0003,
      "step": 780
    },
    {
      "acc": 0.0,
      "epoch": 0.53,
      "learning_rate": 5e-06,
      "loss": 0.1932,
      "rl_loss": 0.0002,
      "step": 790
    },
    {
      "acc": 0.0,
      "epoch": 0.53,
      "learning_rate": 5e-06,
      "loss": 0.2083,
      "rl_loss": 0.0001,
      "step": 800
    },
    {
      "acc": 0.0,
      "epoch": 0.54,
      "learning_rate": 5e-06,
      "loss": 0.1956,
      "rl_loss": 0.0003,
      "step": 810
    },
    {
      "acc": 0.0,
      "epoch": 0.55,
      "learning_rate": 5e-06,
      "loss": 0.1902,
      "rl_loss": 0.0001,
      "step": 820
    },
    {
      "acc": 0.0,
      "epoch": 0.55,
      "learning_rate": 5e-06,
      "loss": 0.1905,
      "rl_loss": 0.0,
      "step": 830
    },
    {
      "acc": 0.0,
      "epoch": 0.56,
      "learning_rate": 5e-06,
      "loss": 0.2264,
      "rl_loss": 0.0002,
      "step": 840
    },
    {
      "acc": 0.0,
      "epoch": 0.57,
      "learning_rate": 5e-06,
      "loss": 0.2166,
      "rl_loss": 0.0001,
      "step": 850
    },
    {
      "acc": 0.0,
      "epoch": 0.57,
      "learning_rate": 5e-06,
      "loss": 0.2009,
      "rl_loss": 0.0001,
      "step": 860
    },
    {
      "acc": 0.0,
      "epoch": 0.58,
      "learning_rate": 5e-06,
      "loss": 0.1867,
      "rl_loss": 0.0003,
      "step": 870
    },
    {
      "acc": 0.0,
      "epoch": 0.59,
      "learning_rate": 5e-06,
      "loss": 0.1964,
      "rl_loss": 0.0002,
      "step": 880
    },
    {
      "acc": 0.0,
      "epoch": 0.59,
      "learning_rate": 5e-06,
      "loss": 0.1889,
      "rl_loss": 0.0001,
      "step": 890
    },
    {
      "acc": 0.0,
      "epoch": 0.6,
      "learning_rate": 5e-06,
      "loss": 0.189,
      "rl_loss": 0.0002,
      "step": 900
    },
    {
      "acc": 0.0,
      "epoch": 0.61,
      "learning_rate": 5e-06,
      "loss": 0.1909,
      "rl_loss": 0.0002,
      "step": 910
    },
    {
      "acc": 0.0,
      "epoch": 0.61,
      "learning_rate": 5e-06,
      "loss": 0.192,
      "rl_loss": 0.0003,
      "step": 920
    },
    {
      "acc": 0.0,
      "epoch": 0.62,
      "learning_rate": 5e-06,
      "loss": 0.1963,
      "rl_loss": 0.0002,
      "step": 930
    },
    {
      "acc": 0.0,
      "epoch": 0.63,
      "learning_rate": 5e-06,
      "loss": 0.1942,
      "rl_loss": 0.0,
      "step": 940
    },
    {
      "acc": 0.0,
      "epoch": 0.63,
      "learning_rate": 5e-06,
      "loss": 0.1938,
      "rl_loss": 0.0002,
      "step": 950
    },
    {
      "acc": 0.0,
      "epoch": 0.64,
      "learning_rate": 5e-06,
      "loss": 0.1922,
      "rl_loss": 0.0002,
      "step": 960
    },
    {
      "acc": 0.0,
      "epoch": 0.65,
      "learning_rate": 5e-06,
      "loss": 0.1937,
      "rl_loss": 0.0001,
      "step": 970
    },
    {
      "acc": 0.0,
      "epoch": 0.65,
      "learning_rate": 5e-06,
      "loss": 0.1914,
      "rl_loss": 0.0003,
      "step": 980
    },
    {
      "acc": 0.0,
      "epoch": 0.66,
      "learning_rate": 5e-06,
      "loss": 0.1988,
      "rl_loss": 0.0,
      "step": 990
    },
    {
      "acc": 0.0,
      "epoch": 0.67,
      "learning_rate": 5e-06,
      "loss": 0.195,
      "rl_loss": 0.0002,
      "step": 1000
    },
    {
      "acc": 0.0,
      "epoch": 0.67,
      "learning_rate": 5e-06,
      "loss": 0.1852,
      "rl_loss": 0.0001,
      "step": 1010
    },
    {
      "acc": 0.0,
      "epoch": 0.68,
      "learning_rate": 5e-06,
      "loss": 0.1956,
      "rl_loss": 0.0004,
      "step": 1020
    },
    {
      "acc": 0.0,
      "epoch": 0.69,
      "learning_rate": 5e-06,
      "loss": 0.2362,
      "rl_loss": 0.0001,
      "step": 1030
    },
    {
      "acc": 0.0,
      "epoch": 0.69,
      "learning_rate": 5e-06,
      "loss": 0.1918,
      "rl_loss": 0.0004,
      "step": 1040
    },
    {
      "acc": 0.0,
      "epoch": 0.7,
      "learning_rate": 5e-06,
      "loss": 0.19,
      "rl_loss": 0.0001,
      "step": 1050
    },
    {
      "acc": 0.0,
      "epoch": 0.71,
      "learning_rate": 5e-06,
      "loss": 0.202,
      "rl_loss": 0.0002,
      "step": 1060
    },
    {
      "acc": 0.0,
      "epoch": 0.71,
      "learning_rate": 5e-06,
      "loss": 0.1978,
      "rl_loss": 0.0001,
      "step": 1070
    },
    {
      "acc": 0.0,
      "epoch": 0.72,
      "learning_rate": 5e-06,
      "loss": 0.2109,
      "rl_loss": 0.0003,
      "step": 1080
    },
    {
      "acc": 0.0,
      "epoch": 0.73,
      "learning_rate": 5e-06,
      "loss": 0.1946,
      "rl_loss": 0.0001,
      "step": 1090
    },
    {
      "acc": 0.0,
      "epoch": 0.73,
      "learning_rate": 5e-06,
      "loss": 0.2027,
      "rl_loss": 0.0001,
      "step": 1100
    },
    {
      "acc": 0.0,
      "epoch": 0.74,
      "learning_rate": 5e-06,
      "loss": 0.1825,
      "rl_loss": 0.0001,
      "step": 1110
    },
    {
      "acc": 0.0,
      "epoch": 0.75,
      "learning_rate": 5e-06,
      "loss": 0.1871,
      "rl_loss": 0.0002,
      "step": 1120
    },
    {
      "acc": 0.0,
      "epoch": 0.75,
      "learning_rate": 5e-06,
      "loss": 0.227,
      "rl_loss": 0.0002,
      "step": 1130
    },
    {
      "acc": 0.0,
      "epoch": 0.76,
      "learning_rate": 5e-06,
      "loss": 0.2044,
      "rl_loss": 0.0002,
      "step": 1140
    },
    {
      "acc": 0.0,
      "epoch": 0.77,
      "learning_rate": 5e-06,
      "loss": 0.2137,
      "rl_loss": 0.0001,
      "step": 1150
    },
    {
      "acc": 0.0,
      "epoch": 0.77,
      "learning_rate": 5e-06,
      "loss": 0.2025,
      "rl_loss": 0.0001,
      "step": 1160
    },
    {
      "acc": 0.0,
      "epoch": 0.78,
      "learning_rate": 5e-06,
      "loss": 0.1859,
      "rl_loss": 0.0001,
      "step": 1170
    },
    {
      "acc": 0.0,
      "epoch": 0.79,
      "learning_rate": 5e-06,
      "loss": 0.1993,
      "rl_loss": 0.0002,
      "step": 1180
    },
    {
      "acc": 0.0,
      "epoch": 0.79,
      "learning_rate": 5e-06,
      "loss": 0.2124,
      "rl_loss": 0.0001,
      "step": 1190
    },
    {
      "acc": 0.0,
      "epoch": 0.8,
      "learning_rate": 5e-06,
      "loss": 0.2046,
      "rl_loss": 0.0003,
      "step": 1200
    },
    {
      "acc": 0.0,
      "epoch": 0.81,
      "learning_rate": 5e-06,
      "loss": 0.1925,
      "rl_loss": 0.0002,
      "step": 1210
    },
    {
      "acc": 0.0,
      "epoch": 0.81,
      "learning_rate": 5e-06,
      "loss": 0.188,
      "rl_loss": 0.0002,
      "step": 1220
    },
    {
      "acc": 0.0,
      "epoch": 0.82,
      "learning_rate": 5e-06,
      "loss": 0.1929,
      "rl_loss": 0.0002,
      "step": 1230
    },
    {
      "acc": 0.0,
      "epoch": 0.83,
      "learning_rate": 5e-06,
      "loss": 0.1991,
      "rl_loss": 0.0001,
      "step": 1240
    },
    {
      "acc": 0.0,
      "epoch": 0.83,
      "learning_rate": 5e-06,
      "loss": 0.1947,
      "rl_loss": 0.0001,
      "step": 1250
    },
    {
      "acc": 0.0,
      "epoch": 0.84,
      "learning_rate": 5e-06,
      "loss": 0.1979,
      "rl_loss": 0.0001,
      "step": 1260
    },
    {
      "acc": 0.0,
      "epoch": 0.85,
      "learning_rate": 5e-06,
      "loss": 0.185,
      "rl_loss": 0.0001,
      "step": 1270
    },
    {
      "acc": 0.0,
      "epoch": 0.85,
      "learning_rate": 5e-06,
      "loss": 0.1845,
      "rl_loss": 0.0001,
      "step": 1280
    },
    {
      "acc": 0.0,
      "epoch": 0.86,
      "learning_rate": 5e-06,
      "loss": 0.1981,
      "rl_loss": 0.0001,
      "step": 1290
    },
    {
      "acc": 0.0,
      "epoch": 0.87,
      "learning_rate": 5e-06,
      "loss": 0.194,
      "rl_loss": 0.0001,
      "step": 1300
    },
    {
      "acc": 0.0,
      "epoch": 0.87,
      "learning_rate": 5e-06,
      "loss": 0.1877,
      "rl_loss": 0.0002,
      "step": 1310
    },
    {
      "acc": 0.0,
      "epoch": 0.88,
      "learning_rate": 5e-06,
      "loss": 0.201,
      "rl_loss": 0.0001,
      "step": 1320
    },
    {
      "acc": 0.0,
      "epoch": 0.89,
      "learning_rate": 5e-06,
      "loss": 0.1941,
      "rl_loss": 0.0004,
      "step": 1330
    },
    {
      "acc": 0.0,
      "epoch": 0.89,
      "learning_rate": 5e-06,
      "loss": 0.1841,
      "rl_loss": 0.0001,
      "step": 1340
    },
    {
      "acc": 0.0,
      "epoch": 0.9,
      "learning_rate": 5e-06,
      "loss": 0.1895,
      "rl_loss": 0.0001,
      "step": 1350
    },
    {
      "acc": 0.0,
      "epoch": 0.91,
      "learning_rate": 5e-06,
      "loss": 0.2057,
      "rl_loss": 0.0002,
      "step": 1360
    },
    {
      "acc": 0.0,
      "epoch": 0.91,
      "learning_rate": 5e-06,
      "loss": 0.1902,
      "rl_loss": 0.0001,
      "step": 1370
    },
    {
      "acc": 0.0,
      "epoch": 0.92,
      "learning_rate": 5e-06,
      "loss": 0.1852,
      "rl_loss": 0.0002,
      "step": 1380
    },
    {
      "acc": 0.0,
      "epoch": 0.93,
      "learning_rate": 5e-06,
      "loss": 0.1908,
      "rl_loss": 0.0002,
      "step": 1390
    },
    {
      "acc": 0.0,
      "epoch": 0.93,
      "learning_rate": 5e-06,
      "loss": 0.1881,
      "rl_loss": 0.0001,
      "step": 1400
    },
    {
      "acc": 0.0,
      "epoch": 0.94,
      "learning_rate": 5e-06,
      "loss": 0.1931,
      "rl_loss": 0.0001,
      "step": 1410
    },
    {
      "acc": 0.0,
      "epoch": 0.95,
      "learning_rate": 5e-06,
      "loss": 0.185,
      "rl_loss": 0.0001,
      "step": 1420
    },
    {
      "acc": 0.0,
      "epoch": 0.95,
      "learning_rate": 5e-06,
      "loss": 0.1873,
      "rl_loss": 0.0002,
      "step": 1430
    },
    {
      "acc": 0.0,
      "epoch": 0.96,
      "learning_rate": 5e-06,
      "loss": 0.1933,
      "rl_loss": 0.0002,
      "step": 1440
    },
    {
      "acc": 0.0,
      "epoch": 0.97,
      "learning_rate": 5e-06,
      "loss": 0.195,
      "rl_loss": 0.0001,
      "step": 1450
    },
    {
      "acc": 0.0,
      "epoch": 0.97,
      "learning_rate": 5e-06,
      "loss": 0.1952,
      "rl_loss": 0.0004,
      "step": 1460
    },
    {
      "acc": 0.0,
      "epoch": 0.98,
      "learning_rate": 5e-06,
      "loss": 0.1894,
      "rl_loss": 0.0002,
      "step": 1470
    },
    {
      "acc": 0.0,
      "epoch": 0.99,
      "learning_rate": 5e-06,
      "loss": 0.191,
      "rl_loss": 0.0001,
      "step": 1480
    },
    {
      "acc": 0.0,
      "epoch": 0.99,
      "learning_rate": 5e-06,
      "loss": 0.188,
      "rl_loss": 0.0004,
      "step": 1490
    },
    {
      "acc": 0.0,
      "epoch": 1.0,
      "learning_rate": 5e-06,
      "loss": 0.1849,
      "rl_loss": 0.0001,
      "step": 1500
    },
    {
      "acc": 0.0,
      "epoch": 1.01,
      "learning_rate": 5e-06,
      "loss": 0.2064,
      "rl_loss": 0.0003,
      "step": 1510
    },
    {
      "acc": 0.0,
      "epoch": 1.01,
      "learning_rate": 5e-06,
      "loss": 0.1824,
      "rl_loss": 0.0002,
      "step": 1520
    },
    {
      "acc": 0.0,
      "epoch": 1.02,
      "learning_rate": 5e-06,
      "loss": 0.1919,
      "rl_loss": 0.0002,
      "step": 1530
    },
    {
      "acc": 0.0,
      "epoch": 1.03,
      "learning_rate": 5e-06,
      "loss": 0.1846,
      "rl_loss": 0.0003,
      "step": 1540
    },
    {
      "acc": 0.0,
      "epoch": 1.03,
      "learning_rate": 5e-06,
      "loss": 0.2003,
      "rl_loss": 0.0001,
      "step": 1550
    },
    {
      "acc": 0.0,
      "epoch": 1.04,
      "learning_rate": 5e-06,
      "loss": 0.1897,
      "rl_loss": 0.0001,
      "step": 1560
    },
    {
      "acc": 0.0,
      "epoch": 1.05,
      "learning_rate": 5e-06,
      "loss": 0.2617,
      "rl_loss": 0.0002,
      "step": 1570
    },
    {
      "acc": 0.0,
      "epoch": 1.05,
      "learning_rate": 5e-06,
      "loss": 0.1938,
      "rl_loss": 0.0001,
      "step": 1580
    },
    {
      "acc": 0.0,
      "epoch": 1.06,
      "learning_rate": 5e-06,
      "loss": 0.1787,
      "rl_loss": 0.0002,
      "step": 1590
    },
    {
      "acc": 0.0,
      "epoch": 1.07,
      "learning_rate": 5e-06,
      "loss": 0.1996,
      "rl_loss": 0.0002,
      "step": 1600
    },
    {
      "acc": 0.0,
      "epoch": 1.07,
      "learning_rate": 5e-06,
      "loss": 0.1918,
      "rl_loss": 0.0004,
      "step": 1610
    },
    {
      "acc": 0.0,
      "epoch": 1.08,
      "learning_rate": 5e-06,
      "loss": 0.2247,
      "rl_loss": 0.0,
      "step": 1620
    },
    {
      "acc": 0.0,
      "epoch": 1.09,
      "learning_rate": 5e-06,
      "loss": 0.1887,
      "rl_loss": 0.0001,
      "step": 1630
    },
    {
      "acc": 0.0,
      "epoch": 1.09,
      "learning_rate": 5e-06,
      "loss": 0.1935,
      "rl_loss": 0.0002,
      "step": 1640
    },
    {
      "acc": 0.0,
      "epoch": 1.1,
      "learning_rate": 5e-06,
      "loss": 0.207,
      "rl_loss": 0.0001,
      "step": 1650
    },
    {
      "acc": 0.0,
      "epoch": 1.11,
      "learning_rate": 5e-06,
      "loss": 0.1844,
      "rl_loss": 0.0002,
      "step": 1660
    },
    {
      "acc": 0.0,
      "epoch": 1.11,
      "learning_rate": 5e-06,
      "loss": 0.2027,
      "rl_loss": 0.0002,
      "step": 1670
    },
    {
      "acc": 0.0,
      "epoch": 1.12,
      "learning_rate": 5e-06,
      "loss": 0.1947,
      "rl_loss": 0.0004,
      "step": 1680
    },
    {
      "acc": 0.0,
      "epoch": 1.13,
      "learning_rate": 5e-06,
      "loss": 0.2074,
      "rl_loss": 0.0001,
      "step": 1690
    },
    {
      "acc": 0.0,
      "epoch": 1.13,
      "learning_rate": 5e-06,
      "loss": 0.1816,
      "rl_loss": 0.0001,
      "step": 1700
    },
    {
      "acc": 0.0,
      "epoch": 1.14,
      "learning_rate": 5e-06,
      "loss": 0.1937,
      "rl_loss": 0.0003,
      "step": 1710
    },
    {
      "acc": 0.0,
      "epoch": 1.15,
      "learning_rate": 5e-06,
      "loss": 0.1973,
      "rl_loss": 0.0002,
      "step": 1720
    },
    {
      "acc": 0.0,
      "epoch": 1.15,
      "learning_rate": 5e-06,
      "loss": 0.1842,
      "rl_loss": 0.0,
      "step": 1730
    },
    {
      "acc": 0.0,
      "epoch": 1.16,
      "learning_rate": 5e-06,
      "loss": 0.1866,
      "rl_loss": 0.0001,
      "step": 1740
    },
    {
      "acc": 0.0,
      "epoch": 1.17,
      "learning_rate": 5e-06,
      "loss": 0.1865,
      "rl_loss": 0.0002,
      "step": 1750
    },
    {
      "acc": 0.0,
      "epoch": 1.17,
      "learning_rate": 5e-06,
      "loss": 0.1936,
      "rl_loss": 0.0001,
      "step": 1760
    },
    {
      "acc": 0.0,
      "epoch": 1.18,
      "learning_rate": 5e-06,
      "loss": 0.1993,
      "rl_loss": 0.0002,
      "step": 1770
    },
    {
      "acc": 0.0,
      "epoch": 1.19,
      "learning_rate": 5e-06,
      "loss": 0.1955,
      "rl_loss": 0.0,
      "step": 1780
    },
    {
      "acc": 0.0,
      "epoch": 1.19,
      "learning_rate": 5e-06,
      "loss": 0.1916,
      "rl_loss": 0.0001,
      "step": 1790
    },
    {
      "acc": 0.0,
      "epoch": 1.2,
      "learning_rate": 5e-06,
      "loss": 0.1913,
      "rl_loss": 0.0001,
      "step": 1800
    },
    {
      "acc": 0.0,
      "epoch": 1.21,
      "learning_rate": 5e-06,
      "loss": 0.2028,
      "rl_loss": 0.0002,
      "step": 1810
    },
    {
      "acc": 0.0,
      "epoch": 1.21,
      "learning_rate": 5e-06,
      "loss": 0.1991,
      "rl_loss": 0.0001,
      "step": 1820
    },
    {
      "acc": 0.0,
      "epoch": 1.22,
      "learning_rate": 5e-06,
      "loss": 0.1758,
      "rl_loss": 0.0003,
      "step": 1830
    },
    {
      "acc": 0.0,
      "epoch": 1.23,
      "learning_rate": 5e-06,
      "loss": 0.1859,
      "rl_loss": 0.0005,
      "step": 1840
    },
    {
      "acc": 0.0,
      "epoch": 1.23,
      "learning_rate": 5e-06,
      "loss": 0.2008,
      "rl_loss": 0.0002,
      "step": 1850
    },
    {
      "acc": 0.0,
      "epoch": 1.24,
      "learning_rate": 5e-06,
      "loss": 0.1778,
      "rl_loss": 0.0001,
      "step": 1860
    },
    {
      "acc": 0.0,
      "epoch": 1.25,
      "learning_rate": 5e-06,
      "loss": 0.1878,
      "rl_loss": 0.0001,
      "step": 1870
    },
    {
      "acc": 0.0,
      "epoch": 1.25,
      "learning_rate": 5e-06,
      "loss": 0.1946,
      "rl_loss": 0.0002,
      "step": 1880
    },
    {
      "acc": 0.0,
      "epoch": 1.26,
      "learning_rate": 5e-06,
      "loss": 0.1814,
      "rl_loss": 0.0001,
      "step": 1890
    },
    {
      "acc": 0.0,
      "epoch": 1.27,
      "learning_rate": 5e-06,
      "loss": 0.1902,
      "rl_loss": 0.0003,
      "step": 1900
    },
    {
      "acc": 0.0,
      "epoch": 1.27,
      "learning_rate": 5e-06,
      "loss": 0.196,
      "rl_loss": 0.0002,
      "step": 1910
    },
    {
      "acc": 0.0,
      "epoch": 1.28,
      "learning_rate": 5e-06,
      "loss": 0.1867,
      "rl_loss": 0.0002,
      "step": 1920
    },
    {
      "acc": 0.0,
      "epoch": 1.29,
      "learning_rate": 5e-06,
      "loss": 0.1947,
      "rl_loss": 0.0002,
      "step": 1930
    },
    {
      "acc": 0.0,
      "epoch": 1.29,
      "learning_rate": 5e-06,
      "loss": 0.1888,
      "rl_loss": 0.0002,
      "step": 1940
    },
    {
      "acc": 0.0,
      "epoch": 1.3,
      "learning_rate": 5e-06,
      "loss": 0.1803,
      "rl_loss": 0.0003,
      "step": 1950
    },
    {
      "acc": 0.0,
      "epoch": 1.31,
      "learning_rate": 5e-06,
      "loss": 0.1947,
      "rl_loss": 0.0002,
      "step": 1960
    },
    {
      "acc": 0.0,
      "epoch": 1.31,
      "learning_rate": 5e-06,
      "loss": 0.1894,
      "rl_loss": 0.0001,
      "step": 1970
    },
    {
      "acc": 0.0,
      "epoch": 1.32,
      "learning_rate": 5e-06,
      "loss": 0.1913,
      "rl_loss": 0.0002,
      "step": 1980
    },
    {
      "acc": 0.0,
      "epoch": 1.33,
      "learning_rate": 5e-06,
      "loss": 0.19,
      "rl_loss": 0.0004,
      "step": 1990
    },
    {
      "acc": 0.0,
      "epoch": 1.33,
      "learning_rate": 5e-06,
      "loss": 0.1861,
      "rl_loss": 0.0001,
      "step": 2000
    },
    {
      "acc": 0.0,
      "epoch": 1.34,
      "learning_rate": 5e-06,
      "loss": 0.1884,
      "rl_loss": 0.0002,
      "step": 2010
    },
    {
      "acc": 0.0,
      "epoch": 1.35,
      "learning_rate": 5e-06,
      "loss": 0.1947,
      "rl_loss": 0.0002,
      "step": 2020
    },
    {
      "acc": 0.0,
      "epoch": 1.35,
      "learning_rate": 5e-06,
      "loss": 0.2011,
      "rl_loss": 0.0002,
      "step": 2030
    },
    {
      "acc": 0.0,
      "epoch": 1.36,
      "learning_rate": 5e-06,
      "loss": 0.1934,
      "rl_loss": 0.0001,
      "step": 2040
    },
    {
      "acc": 0.0,
      "epoch": 1.37,
      "learning_rate": 5e-06,
      "loss": 0.1854,
      "rl_loss": 0.0001,
      "step": 2050
    },
    {
      "acc": 0.0,
      "epoch": 1.37,
      "learning_rate": 5e-06,
      "loss": 0.187,
      "rl_loss": 0.0001,
      "step": 2060
    },
    {
      "acc": 0.0,
      "epoch": 1.38,
      "learning_rate": 5e-06,
      "loss": 0.1879,
      "rl_loss": 0.0002,
      "step": 2070
    },
    {
      "acc": 0.0,
      "epoch": 1.39,
      "learning_rate": 5e-06,
      "loss": 0.1807,
      "rl_loss": 0.0002,
      "step": 2080
    },
    {
      "acc": 0.0,
      "epoch": 1.39,
      "learning_rate": 5e-06,
      "loss": 0.1764,
      "rl_loss": 0.0005,
      "step": 2090
    },
    {
      "acc": 0.0,
      "epoch": 1.4,
      "learning_rate": 5e-06,
      "loss": 0.1943,
      "rl_loss": 0.0002,
      "step": 2100
    },
    {
      "acc": 0.0,
      "epoch": 1.41,
      "learning_rate": 5e-06,
      "loss": 0.1879,
      "rl_loss": 0.0003,
      "step": 2110
    },
    {
      "acc": 0.0,
      "epoch": 1.41,
      "learning_rate": 5e-06,
      "loss": 0.1893,
      "rl_loss": 0.0001,
      "step": 2120
    },
    {
      "acc": 0.0,
      "epoch": 1.42,
      "learning_rate": 5e-06,
      "loss": 0.1869,
      "rl_loss": 0.0004,
      "step": 2130
    },
    {
      "acc": 0.0,
      "epoch": 1.43,
      "learning_rate": 5e-06,
      "loss": 0.1868,
      "rl_loss": 0.0004,
      "step": 2140
    },
    {
      "acc": 0.0,
      "epoch": 1.43,
      "learning_rate": 5e-06,
      "loss": 0.1912,
      "rl_loss": 0.0,
      "step": 2150
    },
    {
      "acc": 0.0,
      "epoch": 1.44,
      "learning_rate": 5e-06,
      "loss": 0.1797,
      "rl_loss": 0.0001,
      "step": 2160
    },
    {
      "acc": 0.0,
      "epoch": 1.45,
      "learning_rate": 5e-06,
      "loss": 0.1994,
      "rl_loss": 0.0002,
      "step": 2170
    },
    {
      "acc": 0.0,
      "epoch": 1.45,
      "learning_rate": 5e-06,
      "loss": 0.1882,
      "rl_loss": 0.0001,
      "step": 2180
    },
    {
      "acc": 0.0,
      "epoch": 1.46,
      "learning_rate": 5e-06,
      "loss": 0.182,
      "rl_loss": 0.0003,
      "step": 2190
    },
    {
      "acc": 0.0,
      "epoch": 1.47,
      "learning_rate": 5e-06,
      "loss": 0.191,
      "rl_loss": 0.0005,
      "step": 2200
    },
    {
      "acc": 0.0,
      "epoch": 1.47,
      "learning_rate": 5e-06,
      "loss": 0.1874,
      "rl_loss": 0.0003,
      "step": 2210
    },
    {
      "acc": 0.0,
      "epoch": 1.48,
      "learning_rate": 5e-06,
      "loss": 0.1856,
      "rl_loss": 0.0002,
      "step": 2220
    },
    {
      "acc": 0.0,
      "epoch": 1.49,
      "learning_rate": 5e-06,
      "loss": 0.1971,
      "rl_loss": 0.0002,
      "step": 2230
    },
    {
      "acc": 0.0,
      "epoch": 1.49,
      "learning_rate": 5e-06,
      "loss": 0.1824,
      "rl_loss": 0.0002,
      "step": 2240
    },
    {
      "acc": 0.0,
      "epoch": 1.5,
      "learning_rate": 5e-06,
      "loss": 0.1793,
      "rl_loss": 0.0003,
      "step": 2250
    },
    {
      "acc": 0.0,
      "epoch": 1.51,
      "learning_rate": 5e-06,
      "loss": 0.1989,
      "rl_loss": 0.0003,
      "step": 2260
    },
    {
      "acc": 0.0,
      "epoch": 1.51,
      "learning_rate": 5e-06,
      "loss": 0.1826,
      "rl_loss": 0.0002,
      "step": 2270
    },
    {
      "acc": 0.0,
      "epoch": 1.52,
      "learning_rate": 5e-06,
      "loss": 0.1907,
      "rl_loss": 0.0001,
      "step": 2280
    },
    {
      "acc": 0.0,
      "epoch": 1.53,
      "learning_rate": 5e-06,
      "loss": 0.1869,
      "rl_loss": 0.0002,
      "step": 2290
    },
    {
      "acc": 0.0,
      "epoch": 1.53,
      "learning_rate": 5e-06,
      "loss": 0.1808,
      "rl_loss": 0.0002,
      "step": 2300
    },
    {
      "acc": 0.0,
      "epoch": 1.54,
      "learning_rate": 5e-06,
      "loss": 0.1965,
      "rl_loss": 0.0003,
      "step": 2310
    },
    {
      "acc": 0.0,
      "epoch": 1.55,
      "learning_rate": 5e-06,
      "loss": 0.1837,
      "rl_loss": 0.0001,
      "step": 2320
    },
    {
      "acc": 0.0,
      "epoch": 1.55,
      "learning_rate": 5e-06,
      "loss": 0.1884,
      "rl_loss": 0.0002,
      "step": 2330
    },
    {
      "acc": 0.0,
      "epoch": 1.56,
      "learning_rate": 5e-06,
      "loss": 0.1869,
      "rl_loss": 0.0004,
      "step": 2340
    },
    {
      "acc": 0.0,
      "epoch": 1.57,
      "learning_rate": 5e-06,
      "loss": 0.1837,
      "rl_loss": 0.0004,
      "step": 2350
    },
    {
      "acc": 0.0,
      "epoch": 1.57,
      "learning_rate": 5e-06,
      "loss": 0.1749,
      "rl_loss": 0.0002,
      "step": 2360
    },
    {
      "acc": 0.0,
      "epoch": 1.58,
      "learning_rate": 5e-06,
      "loss": 0.1775,
      "rl_loss": 0.0002,
      "step": 2370
    },
    {
      "acc": 0.0,
      "epoch": 1.59,
      "learning_rate": 5e-06,
      "loss": 0.178,
      "rl_loss": 0.0002,
      "step": 2380
    },
    {
      "acc": 0.0,
      "epoch": 1.59,
      "learning_rate": 5e-06,
      "loss": 0.1999,
      "rl_loss": 0.0002,
      "step": 2390
    },
    {
      "acc": 0.0,
      "epoch": 1.6,
      "learning_rate": 5e-06,
      "loss": 0.1889,
      "rl_loss": 0.0,
      "step": 2400
    },
    {
      "acc": 0.0,
      "epoch": 1.61,
      "learning_rate": 5e-06,
      "loss": 0.1842,
      "rl_loss": 0.0001,
      "step": 2410
    },
    {
      "acc": 0.0,
      "epoch": 1.61,
      "learning_rate": 5e-06,
      "loss": 0.1888,
      "rl_loss": 0.0001,
      "step": 2420
    },
    {
      "acc": 0.0,
      "epoch": 1.62,
      "learning_rate": 5e-06,
      "loss": 0.1866,
      "rl_loss": 0.0002,
      "step": 2430
    },
    {
      "acc": 0.0,
      "epoch": 1.63,
      "learning_rate": 5e-06,
      "loss": 0.177,
      "rl_loss": 0.0001,
      "step": 2440
    },
    {
      "acc": 0.0,
      "epoch": 1.63,
      "learning_rate": 5e-06,
      "loss": 0.1788,
      "rl_loss": 0.0001,
      "step": 2450
    },
    {
      "acc": 0.0,
      "epoch": 1.64,
      "learning_rate": 5e-06,
      "loss": 0.1858,
      "rl_loss": 0.0001,
      "step": 2460
    },
    {
      "acc": 0.0,
      "epoch": 1.65,
      "learning_rate": 5e-06,
      "loss": 0.1889,
      "rl_loss": 0.0001,
      "step": 2470
    },
    {
      "acc": 0.0,
      "epoch": 1.65,
      "learning_rate": 5e-06,
      "loss": 0.1805,
      "rl_loss": 0.0004,
      "step": 2480
    },
    {
      "acc": 0.0,
      "epoch": 1.66,
      "learning_rate": 5e-06,
      "loss": 0.1895,
      "rl_loss": 0.0001,
      "step": 2490
    },
    {
      "acc": 0.0,
      "epoch": 1.67,
      "learning_rate": 5e-06,
      "loss": 0.1866,
      "rl_loss": 0.0001,
      "step": 2500
    },
    {
      "acc": 0.0,
      "epoch": 1.67,
      "learning_rate": 5e-06,
      "loss": 0.1861,
      "rl_loss": 0.0003,
      "step": 2510
    },
    {
      "acc": 0.0,
      "epoch": 1.68,
      "learning_rate": 5e-06,
      "loss": 0.1792,
      "rl_loss": 0.0004,
      "step": 2520
    },
    {
      "acc": 0.0,
      "epoch": 1.69,
      "learning_rate": 5e-06,
      "loss": 0.1848,
      "rl_loss": 0.0002,
      "step": 2530
    },
    {
      "acc": 0.0,
      "epoch": 1.69,
      "learning_rate": 5e-06,
      "loss": 0.1786,
      "rl_loss": 0.0001,
      "step": 2540
    },
    {
      "acc": 0.0,
      "epoch": 1.7,
      "learning_rate": 5e-06,
      "loss": 0.1789,
      "rl_loss": 0.0,
      "step": 2550
    },
    {
      "acc": 0.0,
      "epoch": 1.71,
      "learning_rate": 5e-06,
      "loss": 0.1841,
      "rl_loss": 0.0,
      "step": 2560
    },
    {
      "acc": 0.0,
      "epoch": 1.71,
      "learning_rate": 5e-06,
      "loss": 0.1841,
      "rl_loss": 0.0001,
      "step": 2570
    },
    {
      "acc": 0.0,
      "epoch": 1.72,
      "learning_rate": 5e-06,
      "loss": 0.1805,
      "rl_loss": 0.0002,
      "step": 2580
    },
    {
      "acc": 0.0,
      "epoch": 1.73,
      "learning_rate": 5e-06,
      "loss": 0.1869,
      "rl_loss": 0.0001,
      "step": 2590
    },
    {
      "acc": 0.0,
      "epoch": 1.73,
      "learning_rate": 5e-06,
      "loss": 0.1873,
      "rl_loss": 0.0002,
      "step": 2600
    },
    {
      "acc": 0.0,
      "epoch": 1.74,
      "learning_rate": 5e-06,
      "loss": 0.1873,
      "rl_loss": 0.0002,
      "step": 2610
    },
    {
      "acc": 0.0,
      "epoch": 1.75,
      "learning_rate": 5e-06,
      "loss": 0.1817,
      "rl_loss": 0.0001,
      "step": 2620
    },
    {
      "acc": 0.0,
      "epoch": 1.75,
      "learning_rate": 5e-06,
      "loss": 0.1821,
      "rl_loss": 0.0003,
      "step": 2630
    },
    {
      "acc": 0.0,
      "epoch": 1.76,
      "learning_rate": 5e-06,
      "loss": 0.1807,
      "rl_loss": 0.0004,
      "step": 2640
    },
    {
      "acc": 0.0,
      "epoch": 1.77,
      "learning_rate": 5e-06,
      "loss": 0.1875,
      "rl_loss": 0.0001,
      "step": 2650
    },
    {
      "acc": 0.0,
      "epoch": 1.77,
      "learning_rate": 5e-06,
      "loss": 0.1788,
      "rl_loss": 0.0002,
      "step": 2660
    },
    {
      "acc": 0.0,
      "epoch": 1.78,
      "learning_rate": 5e-06,
      "loss": 0.1747,
      "rl_loss": 0.0001,
      "step": 2670
    },
    {
      "acc": 0.0,
      "epoch": 1.79,
      "learning_rate": 5e-06,
      "loss": 0.1838,
      "rl_loss": 0.0004,
      "step": 2680
    },
    {
      "acc": 0.0,
      "epoch": 1.79,
      "learning_rate": 5e-06,
      "loss": 0.1942,
      "rl_loss": 0.0003,
      "step": 2690
    },
    {
      "acc": 0.0,
      "epoch": 1.8,
      "learning_rate": 5e-06,
      "loss": 0.1865,
      "rl_loss": 0.0002,
      "step": 2700
    },
    {
      "acc": 0.0,
      "epoch": 1.81,
      "learning_rate": 5e-06,
      "loss": 0.1938,
      "rl_loss": 0.0002,
      "step": 2710
    },
    {
      "acc": 0.0,
      "epoch": 1.81,
      "learning_rate": 5e-06,
      "loss": 0.1865,
      "rl_loss": 0.0002,
      "step": 2720
    },
    {
      "acc": 0.0,
      "epoch": 1.82,
      "learning_rate": 5e-06,
      "loss": 0.1817,
      "rl_loss": 0.0004,
      "step": 2730
    },
    {
      "acc": 0.0,
      "epoch": 1.83,
      "learning_rate": 5e-06,
      "loss": 0.1885,
      "rl_loss": 0.0002,
      "step": 2740
    },
    {
      "acc": 0.0,
      "epoch": 1.83,
      "learning_rate": 5e-06,
      "loss": 0.1842,
      "rl_loss": 0.0003,
      "step": 2750
    },
    {
      "acc": 0.0,
      "epoch": 1.84,
      "learning_rate": 5e-06,
      "loss": 0.1823,
      "rl_loss": 0.0001,
      "step": 2760
    },
    {
      "acc": 0.0,
      "epoch": 1.85,
      "learning_rate": 5e-06,
      "loss": 0.1775,
      "rl_loss": 0.0001,
      "step": 2770
    },
    {
      "acc": 0.0,
      "epoch": 1.85,
      "learning_rate": 5e-06,
      "loss": 0.172,
      "rl_loss": 0.0002,
      "step": 2780
    },
    {
      "acc": 0.0,
      "epoch": 1.86,
      "learning_rate": 5e-06,
      "loss": 0.1903,
      "rl_loss": 0.0002,
      "step": 2790
    },
    {
      "acc": 0.0,
      "epoch": 1.87,
      "learning_rate": 5e-06,
      "loss": 0.1839,
      "rl_loss": 0.0002,
      "step": 2800
    },
    {
      "acc": 0.0,
      "epoch": 1.87,
      "learning_rate": 5e-06,
      "loss": 0.1829,
      "rl_loss": 0.0002,
      "step": 2810
    },
    {
      "acc": 0.0,
      "epoch": 1.88,
      "learning_rate": 5e-06,
      "loss": 0.194,
      "rl_loss": 0.0003,
      "step": 2820
    },
    {
      "acc": 0.0,
      "epoch": 1.89,
      "learning_rate": 5e-06,
      "loss": 0.1933,
      "rl_loss": 0.0001,
      "step": 2830
    },
    {
      "acc": 0.0,
      "epoch": 1.89,
      "learning_rate": 5e-06,
      "loss": 0.1806,
      "rl_loss": 0.0001,
      "step": 2840
    },
    {
      "acc": 0.0,
      "epoch": 1.9,
      "learning_rate": 5e-06,
      "loss": 0.1918,
      "rl_loss": 0.0,
      "step": 2850
    },
    {
      "acc": 0.0,
      "epoch": 1.91,
      "learning_rate": 5e-06,
      "loss": 0.1849,
      "rl_loss": 0.0001,
      "step": 2860
    },
    {
      "acc": 0.0,
      "epoch": 1.91,
      "learning_rate": 5e-06,
      "loss": 0.1833,
      "rl_loss": 0.0001,
      "step": 2870
    },
    {
      "acc": 0.0,
      "epoch": 1.92,
      "learning_rate": 5e-06,
      "loss": 0.191,
      "rl_loss": 0.0001,
      "step": 2880
    },
    {
      "acc": 0.0,
      "epoch": 1.93,
      "learning_rate": 5e-06,
      "loss": 0.1766,
      "rl_loss": 0.0003,
      "step": 2890
    },
    {
      "acc": 0.0,
      "epoch": 1.93,
      "learning_rate": 5e-06,
      "loss": 0.2075,
      "rl_loss": 0.0003,
      "step": 2900
    },
    {
      "acc": 0.0,
      "epoch": 1.94,
      "learning_rate": 5e-06,
      "loss": 0.1822,
      "rl_loss": 0.0002,
      "step": 2910
    },
    {
      "acc": 0.0,
      "epoch": 1.95,
      "learning_rate": 5e-06,
      "loss": 0.1811,
      "rl_loss": 0.0001,
      "step": 2920
    },
    {
      "acc": 0.0,
      "epoch": 1.95,
      "learning_rate": 5e-06,
      "loss": 0.1906,
      "rl_loss": 0.0002,
      "step": 2930
    },
    {
      "acc": 0.0,
      "epoch": 1.96,
      "learning_rate": 5e-06,
      "loss": 0.184,
      "rl_loss": 0.0001,
      "step": 2940
    },
    {
      "acc": 0.0,
      "epoch": 1.97,
      "learning_rate": 5e-06,
      "loss": 0.1876,
      "rl_loss": 0.0002,
      "step": 2950
    },
    {
      "acc": 0.0,
      "epoch": 1.97,
      "learning_rate": 5e-06,
      "loss": 0.1919,
      "rl_loss": 0.0001,
      "step": 2960
    },
    {
      "acc": 0.0,
      "epoch": 1.98,
      "learning_rate": 5e-06,
      "loss": 0.178,
      "rl_loss": 0.0001,
      "step": 2970
    },
    {
      "acc": 0.0,
      "epoch": 1.99,
      "learning_rate": 5e-06,
      "loss": 0.178,
      "rl_loss": 0.0003,
      "step": 2980
    },
    {
      "acc": 0.0,
      "epoch": 1.99,
      "learning_rate": 5e-06,
      "loss": 0.1976,
      "rl_loss": 0.0001,
      "step": 2990
    },
    {
      "acc": 0.0,
      "epoch": 2.0,
      "learning_rate": 5e-06,
      "loss": 0.1775,
      "rl_loss": 0.0002,
      "step": 3000
    },
    {
      "acc": 0.0,
      "epoch": 2.01,
      "learning_rate": 5e-06,
      "loss": 0.2161,
      "rl_loss": 0.0004,
      "step": 3010
    },
    {
      "acc": 0.0,
      "epoch": 2.01,
      "learning_rate": 5e-06,
      "loss": 0.1904,
      "rl_loss": 0.0002,
      "step": 3020
    },
    {
      "acc": 0.0,
      "epoch": 2.02,
      "learning_rate": 5e-06,
      "loss": 0.1829,
      "rl_loss": 0.0002,
      "step": 3030
    },
    {
      "acc": 0.0,
      "epoch": 2.03,
      "learning_rate": 5e-06,
      "loss": 0.184,
      "rl_loss": 0.0002,
      "step": 3040
    },
    {
      "acc": 0.0,
      "epoch": 2.03,
      "learning_rate": 5e-06,
      "loss": 0.1954,
      "rl_loss": 0.0001,
      "step": 3050
    },
    {
      "acc": 0.0,
      "epoch": 2.04,
      "learning_rate": 5e-06,
      "loss": 0.1787,
      "rl_loss": 0.0001,
      "step": 3060
    },
    {
      "acc": 0.0,
      "epoch": 2.05,
      "learning_rate": 5e-06,
      "loss": 0.1874,
      "rl_loss": 0.0002,
      "step": 3070
    },
    {
      "acc": 0.0,
      "epoch": 2.05,
      "learning_rate": 5e-06,
      "loss": 0.1687,
      "rl_loss": 0.0,
      "step": 3080
    },
    {
      "acc": 0.0,
      "epoch": 2.06,
      "learning_rate": 5e-06,
      "loss": 0.2028,
      "rl_loss": 0.0001,
      "step": 3090
    },
    {
      "acc": 0.0,
      "epoch": 2.07,
      "learning_rate": 5e-06,
      "loss": 0.1778,
      "rl_loss": 0.0002,
      "step": 3100
    },
    {
      "acc": 0.0,
      "epoch": 2.07,
      "learning_rate": 5e-06,
      "loss": 0.1854,
      "rl_loss": 0.0002,
      "step": 3110
    },
    {
      "acc": 0.0,
      "epoch": 2.08,
      "learning_rate": 5e-06,
      "loss": 0.189,
      "rl_loss": 0.0004,
      "step": 3120
    },
    {
      "acc": 0.0,
      "epoch": 2.09,
      "learning_rate": 5e-06,
      "loss": 0.1775,
      "rl_loss": 0.0002,
      "step": 3130
    },
    {
      "acc": 0.0,
      "epoch": 2.09,
      "learning_rate": 5e-06,
      "loss": 0.1764,
      "rl_loss": 0.0002,
      "step": 3140
    },
    {
      "acc": 0.0,
      "epoch": 2.1,
      "learning_rate": 5e-06,
      "loss": 0.196,
      "rl_loss": 0.0004,
      "step": 3150
    },
    {
      "acc": 0.0,
      "epoch": 2.11,
      "learning_rate": 5e-06,
      "loss": 0.1824,
      "rl_loss": 0.0004,
      "step": 3160
    },
    {
      "acc": 0.0,
      "epoch": 2.11,
      "learning_rate": 5e-06,
      "loss": 0.1883,
      "rl_loss": 0.0001,
      "step": 3170
    },
    {
      "acc": 0.0,
      "epoch": 2.12,
      "learning_rate": 5e-06,
      "loss": 0.1837,
      "rl_loss": 0.0002,
      "step": 3180
    },
    {
      "acc": 0.0,
      "epoch": 2.13,
      "learning_rate": 5e-06,
      "loss": 0.1897,
      "rl_loss": 0.0002,
      "step": 3190
    },
    {
      "acc": 0.0,
      "epoch": 2.13,
      "learning_rate": 5e-06,
      "loss": 0.1858,
      "rl_loss": 0.0001,
      "step": 3200
    },
    {
      "acc": 0.0,
      "epoch": 2.14,
      "learning_rate": 5e-06,
      "loss": 0.1867,
      "rl_loss": 0.0001,
      "step": 3210
    },
    {
      "acc": 0.0,
      "epoch": 2.15,
      "learning_rate": 5e-06,
      "loss": 0.1856,
      "rl_loss": 0.0001,
      "step": 3220
    },
    {
      "acc": 0.0,
      "epoch": 2.15,
      "learning_rate": 5e-06,
      "loss": 0.1777,
      "rl_loss": 0.0001,
      "step": 3230
    },
    {
      "acc": 0.0,
      "epoch": 2.16,
      "learning_rate": 5e-06,
      "loss": 0.1848,
      "rl_loss": 0.0001,
      "step": 3240
    },
    {
      "acc": 0.0,
      "epoch": 2.17,
      "learning_rate": 5e-06,
      "loss": 0.183,
      "rl_loss": 0.0001,
      "step": 3250
    },
    {
      "acc": 0.0,
      "epoch": 2.17,
      "learning_rate": 5e-06,
      "loss": 0.1887,
      "rl_loss": 0.0001,
      "step": 3260
    },
    {
      "acc": 0.0,
      "epoch": 2.18,
      "learning_rate": 5e-06,
      "loss": 0.1789,
      "rl_loss": 0.0001,
      "step": 3270
    },
    {
      "acc": 0.0,
      "epoch": 2.19,
      "learning_rate": 5e-06,
      "loss": 0.1798,
      "rl_loss": 0.0003,
      "step": 3280
    },
    {
      "acc": 0.0,
      "epoch": 2.19,
      "learning_rate": 5e-06,
      "loss": 0.181,
      "rl_loss": 0.0,
      "step": 3290
    },
    {
      "acc": 0.0,
      "epoch": 2.2,
      "learning_rate": 5e-06,
      "loss": 0.1769,
      "rl_loss": 0.0002,
      "step": 3300
    },
    {
      "acc": 0.0,
      "epoch": 2.21,
      "learning_rate": 5e-06,
      "loss": 0.192,
      "rl_loss": 0.0001,
      "step": 3310
    },
    {
      "acc": 0.0,
      "epoch": 2.21,
      "learning_rate": 5e-06,
      "loss": 0.1769,
      "rl_loss": 0.0003,
      "step": 3320
    },
    {
      "acc": 0.0,
      "epoch": 2.22,
      "learning_rate": 5e-06,
      "loss": 0.1926,
      "rl_loss": 0.0002,
      "step": 3330
    },
    {
      "acc": 0.0,
      "epoch": 2.23,
      "learning_rate": 5e-06,
      "loss": 0.1977,
      "rl_loss": 0.0003,
      "step": 3340
    },
    {
      "acc": 0.0,
      "epoch": 2.23,
      "learning_rate": 5e-06,
      "loss": 0.1823,
      "rl_loss": 0.0003,
      "step": 3350
    },
    {
      "acc": 0.0,
      "epoch": 2.24,
      "learning_rate": 5e-06,
      "loss": 0.1906,
      "rl_loss": 0.0001,
      "step": 3360
    },
    {
      "acc": 0.0,
      "epoch": 2.25,
      "learning_rate": 5e-06,
      "loss": 0.1821,
      "rl_loss": 0.0003,
      "step": 3370
    },
    {
      "acc": 0.0,
      "epoch": 2.25,
      "learning_rate": 5e-06,
      "loss": 0.1862,
      "rl_loss": 0.0002,
      "step": 3380
    },
    {
      "acc": 0.0,
      "epoch": 2.26,
      "learning_rate": 5e-06,
      "loss": 0.187,
      "rl_loss": 0.0001,
      "step": 3390
    },
    {
      "acc": 0.0,
      "epoch": 2.27,
      "learning_rate": 5e-06,
      "loss": 0.1823,
      "rl_loss": 0.0,
      "step": 3400
    },
    {
      "acc": 0.0,
      "epoch": 2.27,
      "learning_rate": 5e-06,
      "loss": 0.1822,
      "rl_loss": 0.0001,
      "step": 3410
    },
    {
      "acc": 0.0,
      "epoch": 2.28,
      "learning_rate": 5e-06,
      "loss": 0.1799,
      "rl_loss": 0.0001,
      "step": 3420
    },
    {
      "acc": 0.0,
      "epoch": 2.29,
      "learning_rate": 5e-06,
      "loss": 0.1868,
      "rl_loss": 0.0004,
      "step": 3430
    },
    {
      "acc": 0.0,
      "epoch": 2.29,
      "learning_rate": 5e-06,
      "loss": 0.187,
      "rl_loss": 0.0001,
      "step": 3440
    },
    {
      "acc": 0.0,
      "epoch": 2.3,
      "learning_rate": 5e-06,
      "loss": 0.1864,
      "rl_loss": 0.0,
      "step": 3450
    },
    {
      "acc": 0.0,
      "epoch": 2.31,
      "learning_rate": 5e-06,
      "loss": 0.1773,
      "rl_loss": 0.0001,
      "step": 3460
    },
    {
      "acc": 0.0,
      "epoch": 2.31,
      "learning_rate": 5e-06,
      "loss": 0.1844,
      "rl_loss": 0.0001,
      "step": 3470
    },
    {
      "acc": 0.0,
      "epoch": 2.32,
      "learning_rate": 5e-06,
      "loss": 0.1831,
      "rl_loss": 0.0002,
      "step": 3480
    },
    {
      "acc": 0.0,
      "epoch": 2.33,
      "learning_rate": 5e-06,
      "loss": 0.1916,
      "rl_loss": 0.0002,
      "step": 3490
    },
    {
      "acc": 0.0,
      "epoch": 2.33,
      "learning_rate": 5e-06,
      "loss": 0.1724,
      "rl_loss": 0.0002,
      "step": 3500
    },
    {
      "acc": 0.0,
      "epoch": 2.34,
      "learning_rate": 5e-06,
      "loss": 0.1741,
      "rl_loss": 0.0002,
      "step": 3510
    },
    {
      "acc": 0.0,
      "epoch": 2.35,
      "learning_rate": 5e-06,
      "loss": 0.1874,
      "rl_loss": 0.0001,
      "step": 3520
    },
    {
      "acc": 0.0,
      "epoch": 2.35,
      "learning_rate": 5e-06,
      "loss": 0.189,
      "rl_loss": 0.0002,
      "step": 3530
    },
    {
      "acc": 0.0,
      "epoch": 2.36,
      "learning_rate": 5e-06,
      "loss": 0.1885,
      "rl_loss": 0.0002,
      "step": 3540
    },
    {
      "acc": 0.0,
      "epoch": 2.37,
      "learning_rate": 5e-06,
      "loss": 0.1846,
      "rl_loss": 0.0001,
      "step": 3550
    },
    {
      "acc": 0.0,
      "epoch": 2.37,
      "learning_rate": 5e-06,
      "loss": 0.1789,
      "rl_loss": 0.0004,
      "step": 3560
    },
    {
      "acc": 0.0,
      "epoch": 2.38,
      "learning_rate": 5e-06,
      "loss": 0.1762,
      "rl_loss": 0.0003,
      "step": 3570
    },
    {
      "acc": 0.0,
      "epoch": 2.39,
      "learning_rate": 5e-06,
      "loss": 0.1799,
      "rl_loss": 0.0,
      "step": 3580
    },
    {
      "acc": 0.0,
      "epoch": 2.39,
      "learning_rate": 5e-06,
      "loss": 0.1876,
      "rl_loss": 0.0006,
      "step": 3590
    },
    {
      "acc": 0.0,
      "epoch": 2.4,
      "learning_rate": 5e-06,
      "loss": 0.1736,
      "rl_loss": 0.0003,
      "step": 3600
    },
    {
      "acc": 0.0,
      "epoch": 2.41,
      "learning_rate": 5e-06,
      "loss": 0.1941,
      "rl_loss": 0.0001,
      "step": 3610
    },
    {
      "acc": 0.0,
      "epoch": 2.41,
      "learning_rate": 5e-06,
      "loss": 0.203,
      "rl_loss": 0.0002,
      "step": 3620
    },
    {
      "acc": 0.0,
      "epoch": 2.42,
      "learning_rate": 5e-06,
      "loss": 0.1885,
      "rl_loss": 0.0001,
      "step": 3630
    },
    {
      "acc": 0.0,
      "epoch": 2.43,
      "learning_rate": 5e-06,
      "loss": 0.1854,
      "rl_loss": 0.0001,
      "step": 3640
    },
    {
      "acc": 0.0,
      "epoch": 2.43,
      "learning_rate": 5e-06,
      "loss": 0.189,
      "rl_loss": 0.0002,
      "step": 3650
    },
    {
      "acc": 0.0,
      "epoch": 2.44,
      "learning_rate": 5e-06,
      "loss": 0.1722,
      "rl_loss": 0.0002,
      "step": 3660
    },
    {
      "acc": 0.0,
      "epoch": 2.45,
      "learning_rate": 5e-06,
      "loss": 0.1849,
      "rl_loss": 0.0001,
      "step": 3670
    },
    {
      "acc": 0.0,
      "epoch": 2.45,
      "learning_rate": 5e-06,
      "loss": 0.1816,
      "rl_loss": 0.0004,
      "step": 3680
    },
    {
      "acc": 0.0,
      "epoch": 2.46,
      "learning_rate": 5e-06,
      "loss": 0.1818,
      "rl_loss": 0.0002,
      "step": 3690
    },
    {
      "acc": 0.0,
      "epoch": 2.47,
      "learning_rate": 5e-06,
      "loss": 0.1749,
      "rl_loss": 0.0001,
      "step": 3700
    },
    {
      "acc": 0.0,
      "epoch": 2.47,
      "learning_rate": 5e-06,
      "loss": 0.1796,
      "rl_loss": 0.0,
      "step": 3710
    },
    {
      "acc": 0.0,
      "epoch": 2.48,
      "learning_rate": 5e-06,
      "loss": 0.1794,
      "rl_loss": 0.0002,
      "step": 3720
    },
    {
      "acc": 0.0,
      "epoch": 2.49,
      "learning_rate": 5e-06,
      "loss": 0.1813,
      "rl_loss": 0.0001,
      "step": 3730
    },
    {
      "acc": 0.0,
      "epoch": 2.49,
      "learning_rate": 5e-06,
      "loss": 0.1818,
      "rl_loss": 0.0001,
      "step": 3740
    },
    {
      "acc": 0.0,
      "epoch": 2.5,
      "learning_rate": 5e-06,
      "loss": 0.1899,
      "rl_loss": 0.0001,
      "step": 3750
    },
    {
      "acc": 0.0,
      "epoch": 2.51,
      "learning_rate": 5e-06,
      "loss": 0.1834,
      "rl_loss": 0.0,
      "step": 3760
    },
    {
      "acc": 0.0,
      "epoch": 2.51,
      "learning_rate": 5e-06,
      "loss": 0.1769,
      "rl_loss": 0.0004,
      "step": 3770
    },
    {
      "acc": 0.0,
      "epoch": 2.52,
      "learning_rate": 5e-06,
      "loss": 0.1771,
      "rl_loss": 0.0001,
      "step": 3780
    },
    {
      "acc": 0.0,
      "epoch": 2.53,
      "learning_rate": 5e-06,
      "loss": 0.1778,
      "rl_loss": 0.0004,
      "step": 3790
    },
    {
      "acc": 0.0,
      "epoch": 2.53,
      "learning_rate": 5e-06,
      "loss": 0.1833,
      "rl_loss": 0.0001,
      "step": 3800
    },
    {
      "acc": 0.0,
      "epoch": 2.54,
      "learning_rate": 5e-06,
      "loss": 0.174,
      "rl_loss": 0.0006,
      "step": 3810
    },
    {
      "acc": 0.0,
      "epoch": 2.55,
      "learning_rate": 5e-06,
      "loss": 0.1743,
      "rl_loss": 0.0001,
      "step": 3820
    },
    {
      "acc": 0.0,
      "epoch": 2.55,
      "learning_rate": 5e-06,
      "loss": 0.1955,
      "rl_loss": 0.0001,
      "step": 3830
    },
    {
      "acc": 0.0,
      "epoch": 2.56,
      "learning_rate": 5e-06,
      "loss": 0.1876,
      "rl_loss": 0.0001,
      "step": 3840
    },
    {
      "acc": 0.0,
      "epoch": 2.57,
      "learning_rate": 5e-06,
      "loss": 0.1868,
      "rl_loss": 0.0002,
      "step": 3850
    },
    {
      "acc": 0.0,
      "epoch": 2.57,
      "learning_rate": 5e-06,
      "loss": 0.1835,
      "rl_loss": 0.0001,
      "step": 3860
    },
    {
      "acc": 0.0,
      "epoch": 2.58,
      "learning_rate": 5e-06,
      "loss": 0.1851,
      "rl_loss": 0.0002,
      "step": 3870
    },
    {
      "acc": 0.0,
      "epoch": 2.59,
      "learning_rate": 5e-06,
      "loss": 0.1795,
      "rl_loss": 0.0003,
      "step": 3880
    },
    {
      "acc": 0.0,
      "epoch": 2.59,
      "learning_rate": 5e-06,
      "loss": 0.1934,
      "rl_loss": 0.0001,
      "step": 3890
    },
    {
      "acc": 0.0,
      "epoch": 2.6,
      "learning_rate": 5e-06,
      "loss": 0.1781,
      "rl_loss": 0.0001,
      "step": 3900
    },
    {
      "acc": 0.0,
      "epoch": 2.61,
      "learning_rate": 5e-06,
      "loss": 0.1667,
      "rl_loss": 0.0001,
      "step": 3910
    },
    {
      "acc": 0.0,
      "epoch": 2.61,
      "learning_rate": 5e-06,
      "loss": 0.1717,
      "rl_loss": 0.0001,
      "step": 3920
    },
    {
      "acc": 0.0,
      "epoch": 2.62,
      "learning_rate": 5e-06,
      "loss": 0.1775,
      "rl_loss": 0.0003,
      "step": 3930
    },
    {
      "acc": 0.0,
      "epoch": 2.63,
      "learning_rate": 5e-06,
      "loss": 0.1871,
      "rl_loss": 0.0003,
      "step": 3940
    },
    {
      "acc": 0.0,
      "epoch": 2.63,
      "learning_rate": 5e-06,
      "loss": 0.177,
      "rl_loss": 0.0003,
      "step": 3950
    },
    {
      "acc": 0.0,
      "epoch": 2.64,
      "learning_rate": 5e-06,
      "loss": 0.1855,
      "rl_loss": 0.0002,
      "step": 3960
    },
    {
      "acc": 0.0,
      "epoch": 2.65,
      "learning_rate": 5e-06,
      "loss": 0.1911,
      "rl_loss": 0.0002,
      "step": 3970
    },
    {
      "acc": 0.0,
      "epoch": 2.65,
      "learning_rate": 5e-06,
      "loss": 0.1796,
      "rl_loss": 0.0002,
      "step": 3980
    },
    {
      "acc": 0.0,
      "epoch": 2.66,
      "learning_rate": 5e-06,
      "loss": 0.1927,
      "rl_loss": 0.0001,
      "step": 3990
    },
    {
      "acc": 0.0,
      "epoch": 2.67,
      "learning_rate": 5e-06,
      "loss": 0.1828,
      "rl_loss": 0.0001,
      "step": 4000
    }
  ],
  "max_steps": 4500,
  "num_train_epochs": 3,
  "total_flos": 6.49818648576e+17,
  "trial_name": null,
  "trial_params": null
}