{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.4515819023292278,
  "global_step": 1793,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 8.228161798644422e-06,
      "loss": 3.1603,
      "step": 5
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.1771838201355582e-05,
      "loss": 2.7601,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 1.384475601163205e-05,
      "loss": 2.6729,
      "step": 15
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.5315514604066738e-05,
      "loss": 2.3958,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.6456323597288844e-05,
      "loss": 2.3595,
      "step": 25
    },
    {
      "epoch": 0.02,
      "learning_rate": 1.738843241434321e-05,
      "loss": 2.4,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.817651918997498e-05,
      "loss": 2.2831,
      "step": 35
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.8859191006777896e-05,
      "loss": 2.2999,
      "step": 40
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.946135022461968e-05,
      "loss": 2.1469,
      "step": 45
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 2.2508,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 2e-05,
      "loss": 2.285,
      "step": 55
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 2.2009,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 2.1514,
      "step": 65
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 2.137,
      "step": 70
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 2.2193,
      "step": 75
    },
    {
      "epoch": 0.06,
      "learning_rate": 2e-05,
      "loss": 2.2193,
      "step": 80
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 2.1647,
      "step": 85
    },
    {
      "epoch": 0.07,
      "learning_rate": 2e-05,
      "loss": 2.2805,
      "step": 90
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 2.1456,
      "step": 95
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 2.1043,
      "step": 100
    },
    {
      "epoch": 0.08,
      "learning_rate": 2e-05,
      "loss": 2.1215,
      "step": 105
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 2.1528,
      "step": 110
    },
    {
      "epoch": 0.09,
      "learning_rate": 2e-05,
      "loss": 2.134,
      "step": 115
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 2.0645,
      "step": 120
    },
    {
      "epoch": 0.1,
      "learning_rate": 2e-05,
      "loss": 2.1347,
      "step": 125
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 2.1372,
      "step": 130
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 2.0921,
      "step": 135
    },
    {
      "epoch": 0.11,
      "learning_rate": 2e-05,
      "loss": 2.1157,
      "step": 140
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 2.0916,
      "step": 145
    },
    {
      "epoch": 0.12,
      "learning_rate": 2e-05,
      "loss": 2.0519,
      "step": 150
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 2.0987,
      "step": 155
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 1.9597,
      "step": 160
    },
    {
      "epoch": 0.13,
      "learning_rate": 2e-05,
      "loss": 2.1301,
      "step": 165
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 2.1803,
      "step": 170
    },
    {
      "epoch": 0.14,
      "learning_rate": 2e-05,
      "loss": 2.1488,
      "step": 175
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 2.0421,
      "step": 180
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 2.2094,
      "step": 185
    },
    {
      "epoch": 0.15,
      "learning_rate": 2e-05,
      "loss": 2.1173,
      "step": 190
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 2.0779,
      "step": 195
    },
    {
      "epoch": 0.16,
      "learning_rate": 2e-05,
      "loss": 2.1213,
      "step": 200
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 2.1425,
      "step": 205
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 2.0807,
      "step": 210
    },
    {
      "epoch": 0.17,
      "learning_rate": 2e-05,
      "loss": 2.1772,
      "step": 215
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 2.0721,
      "step": 220
    },
    {
      "epoch": 0.18,
      "learning_rate": 2e-05,
      "loss": 2.1006,
      "step": 225
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 2.1579,
      "step": 230
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 1.9796,
      "step": 235
    },
    {
      "epoch": 0.19,
      "learning_rate": 2e-05,
      "loss": 2.0889,
      "step": 240
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 1.977,
      "step": 245
    },
    {
      "epoch": 0.2,
      "learning_rate": 2e-05,
      "loss": 2.1304,
      "step": 250
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 2.1525,
      "step": 255
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 2.0773,
      "step": 260
    },
    {
      "epoch": 0.21,
      "learning_rate": 2e-05,
      "loss": 2.145,
      "step": 265
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 2.0542,
      "step": 270
    },
    {
      "epoch": 0.22,
      "learning_rate": 2e-05,
      "loss": 2.0588,
      "step": 275
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 2.2034,
      "step": 280
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 2.0698,
      "step": 285
    },
    {
      "epoch": 0.23,
      "learning_rate": 2e-05,
      "loss": 2.0572,
      "step": 290
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 2.1129,
      "step": 295
    },
    {
      "epoch": 0.24,
      "learning_rate": 2e-05,
      "loss": 2.1429,
      "step": 300
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 2.0936,
      "step": 305
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 1.9964,
      "step": 310
    },
    {
      "epoch": 0.25,
      "learning_rate": 2e-05,
      "loss": 2.1298,
      "step": 315
    },
    {
      "epoch": 0.26,
      "learning_rate": 2e-05,
      "loss": 2.0689,
      "step": 320
    },
    {
      "epoch": 0.26,
      "learning_rate": 2e-05,
      "loss": 2.0804,
      "step": 325
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 2.0465,
      "step": 330
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 2.0871,
      "step": 335
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 2.099,
      "step": 340
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 2.0433,
      "step": 345
    },
    {
      "epoch": 0.28,
      "learning_rate": 2e-05,
      "loss": 2.0825,
      "step": 350
    },
    {
      "epoch": 0.29,
      "learning_rate": 2e-05,
      "loss": 2.0902,
      "step": 355
    },
    {
      "epoch": 0.29,
      "learning_rate": 2e-05,
      "loss": 2.063,
      "step": 360
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 2.1341,
      "step": 365
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 2.0562,
      "step": 370
    },
    {
      "epoch": 0.3,
      "learning_rate": 2e-05,
      "loss": 2.0492,
      "step": 375
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 1.9866,
      "step": 380
    },
    {
      "epoch": 0.31,
      "learning_rate": 2e-05,
      "loss": 2.1307,
      "step": 385
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 2.1265,
      "step": 390
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 2.0729,
      "step": 395
    },
    {
      "epoch": 0.32,
      "learning_rate": 2e-05,
      "loss": 2.0072,
      "step": 400
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 2.1163,
      "step": 405
    },
    {
      "epoch": 0.33,
      "learning_rate": 2e-05,
      "loss": 2.1249,
      "step": 410
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 2.0166,
      "step": 415
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 2.0386,
      "step": 420
    },
    {
      "epoch": 0.34,
      "learning_rate": 2e-05,
      "loss": 2.0629,
      "step": 425
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 2.0647,
      "step": 430
    },
    {
      "epoch": 0.35,
      "learning_rate": 2e-05,
      "loss": 2.1011,
      "step": 435
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 1.9827,
      "step": 440
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 2.1154,
      "step": 445
    },
    {
      "epoch": 0.36,
      "learning_rate": 2e-05,
      "loss": 2.0073,
      "step": 450
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 2.11,
      "step": 455
    },
    {
      "epoch": 0.37,
      "learning_rate": 2e-05,
      "loss": 2.0181,
      "step": 460
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 2.0479,
      "step": 465
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 2.0791,
      "step": 470
    },
    {
      "epoch": 0.38,
      "learning_rate": 2e-05,
      "loss": 2.1109,
      "step": 475
    },
    {
      "epoch": 0.39,
      "learning_rate": 2e-05,
      "loss": 2.0062,
      "step": 480
    },
    {
      "epoch": 0.39,
      "learning_rate": 2e-05,
      "loss": 2.0485,
      "step": 485
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 1.9999,
      "step": 490
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 1.9982,
      "step": 495
    },
    {
      "epoch": 0.4,
      "learning_rate": 2e-05,
      "loss": 2.0399,
      "step": 500
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 2.0282,
      "step": 505
    },
    {
      "epoch": 0.41,
      "learning_rate": 2e-05,
      "loss": 2.064,
      "step": 510
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 2.0239,
      "step": 515
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 1.9875,
      "step": 520
    },
    {
      "epoch": 0.42,
      "learning_rate": 2e-05,
      "loss": 2.1219,
      "step": 525
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 1.9841,
      "step": 530
    },
    {
      "epoch": 0.43,
      "learning_rate": 2e-05,
      "loss": 2.0484,
      "step": 535
    },
    {
      "epoch": 0.44,
      "learning_rate": 2e-05,
      "loss": 2.0757,
      "step": 540
    },
    {
      "epoch": 0.44,
      "learning_rate": 2e-05,
      "loss": 2.0275,
      "step": 545
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 2.1235,
      "step": 550
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 1.994,
      "step": 555
    },
    {
      "epoch": 0.45,
      "learning_rate": 2e-05,
      "loss": 2.0548,
      "step": 560
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 2.0246,
      "step": 565
    },
    {
      "epoch": 0.46,
      "learning_rate": 2e-05,
      "loss": 2.0438,
      "step": 570
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 1.9877,
      "step": 575
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 2.0073,
      "step": 580
    },
    {
      "epoch": 0.47,
      "learning_rate": 2e-05,
      "loss": 2.0335,
      "step": 585
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 1.9349,
      "step": 590
    },
    {
      "epoch": 0.48,
      "learning_rate": 2e-05,
      "loss": 2.0598,
      "step": 595
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 2.1281,
      "step": 600
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 2.1086,
      "step": 605
    },
    {
      "epoch": 0.49,
      "learning_rate": 2e-05,
      "loss": 2.0188,
      "step": 610
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 2.0448,
      "step": 615
    },
    {
      "epoch": 0.5,
      "learning_rate": 2e-05,
      "loss": 2.1233,
      "step": 620
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 1.9751,
      "step": 625
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 2.0676,
      "step": 630
    },
    {
      "epoch": 0.51,
      "learning_rate": 2e-05,
      "loss": 2.0068,
      "step": 635
    },
    {
      "epoch": 0.52,
      "learning_rate": 2e-05,
      "loss": 2.0636,
      "step": 640
    },
    {
      "epoch": 0.52,
      "learning_rate": 2e-05,
      "loss": 2.012,
      "step": 645
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 2.0323,
      "step": 650
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 2.0697,
      "step": 655
    },
    {
      "epoch": 0.53,
      "learning_rate": 2e-05,
      "loss": 2.0221,
      "step": 660
    },
    {
      "epoch": 0.54,
      "learning_rate": 2e-05,
      "loss": 2.0391,
      "step": 665
    },
    {
      "epoch": 0.54,
      "learning_rate": 2e-05,
      "loss": 2.0517,
      "step": 670
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 2.1092,
      "step": 675
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 2.0735,
      "step": 680
    },
    {
      "epoch": 0.55,
      "learning_rate": 2e-05,
      "loss": 2.1768,
      "step": 685
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 1.9814,
      "step": 690
    },
    {
      "epoch": 0.56,
      "learning_rate": 2e-05,
      "loss": 2.0082,
      "step": 695
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 1.9932,
      "step": 700
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 2.0701,
      "step": 705
    },
    {
      "epoch": 0.57,
      "learning_rate": 2e-05,
      "loss": 1.9849,
      "step": 710
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 2.0469,
      "step": 715
    },
    {
      "epoch": 0.58,
      "learning_rate": 2e-05,
      "loss": 2.0328,
      "step": 720
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 2.1638,
      "step": 725
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 1.9913,
      "step": 730
    },
    {
      "epoch": 0.59,
      "learning_rate": 2e-05,
      "loss": 2.0312,
      "step": 735
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 2.0468,
      "step": 740
    },
    {
      "epoch": 0.6,
      "learning_rate": 2e-05,
      "loss": 1.9856,
      "step": 745
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 2.0258,
      "step": 750
    },
    {
      "epoch": 0.61,
      "learning_rate": 2e-05,
      "loss": 2.0983,
      "step": 755
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 1.9871,
      "step": 760
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 2.01,
      "step": 765
    },
    {
      "epoch": 0.62,
      "learning_rate": 2e-05,
      "loss": 2.0858,
      "step": 770
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 2.056,
      "step": 775
    },
    {
      "epoch": 0.63,
      "learning_rate": 2e-05,
      "loss": 2.0671,
      "step": 780
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 2.0151,
      "step": 785
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 2.014,
      "step": 790
    },
    {
      "epoch": 0.64,
      "learning_rate": 2e-05,
      "loss": 1.9757,
      "step": 795
    },
    {
      "epoch": 0.65,
      "learning_rate": 2e-05,
      "loss": 2.0435,
      "step": 800
    },
    {
      "epoch": 0.65,
      "learning_rate": 2e-05,
      "loss": 1.9699,
      "step": 805
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 2.1073,
      "step": 810
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 2.0371,
      "step": 815
    },
    {
      "epoch": 0.66,
      "learning_rate": 2e-05,
      "loss": 1.9612,
      "step": 820
    },
    {
      "epoch": 0.67,
      "learning_rate": 2e-05,
      "loss": 2.0357,
      "step": 825
    },
    {
      "epoch": 0.67,
      "learning_rate": 2e-05,
      "loss": 1.9747,
      "step": 830
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 2.0736,
      "step": 835
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 2.0831,
      "step": 840
    },
    {
      "epoch": 0.68,
      "learning_rate": 2e-05,
      "loss": 2.1133,
      "step": 845
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 1.9934,
      "step": 850
    },
    {
      "epoch": 0.69,
      "learning_rate": 2e-05,
      "loss": 1.8631,
      "step": 855
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 2.0426,
      "step": 860
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 2.0171,
      "step": 865
    },
    {
      "epoch": 0.7,
      "learning_rate": 2e-05,
      "loss": 1.9415,
      "step": 870
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 1.914,
      "step": 875
    },
    {
      "epoch": 0.71,
      "learning_rate": 2e-05,
      "loss": 2.0661,
      "step": 880
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 2.0792,
      "step": 885
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 1.9541,
      "step": 890
    },
    {
      "epoch": 0.72,
      "learning_rate": 2e-05,
      "loss": 2.0045,
      "step": 895
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 2.0326,
      "step": 900
    },
    {
      "epoch": 0.73,
      "learning_rate": 2e-05,
      "loss": 2.0406,
      "step": 905
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 2.0131,
      "step": 910
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 2.0479,
      "step": 915
    },
    {
      "epoch": 0.74,
      "learning_rate": 2e-05,
      "loss": 1.9801,
      "step": 920
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 2.0129,
      "step": 925
    },
    {
      "epoch": 0.75,
      "learning_rate": 2e-05,
      "loss": 1.9463,
      "step": 930
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 1.9324,
      "step": 935
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 1.9729,
      "step": 940
    },
    {
      "epoch": 0.76,
      "learning_rate": 2e-05,
      "loss": 2.1066,
      "step": 945
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 1.9615,
      "step": 950
    },
    {
      "epoch": 0.77,
      "learning_rate": 2e-05,
      "loss": 2.054,
      "step": 955
    },
    {
      "epoch": 0.78,
      "learning_rate": 2e-05,
      "loss": 1.9475,
      "step": 960
    },
    {
      "epoch": 0.78,
      "learning_rate": 2e-05,
      "loss": 1.9504,
      "step": 965
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 1.9985,
      "step": 970
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 1.9808,
      "step": 975
    },
    {
      "epoch": 0.79,
      "learning_rate": 2e-05,
      "loss": 1.9699,
      "step": 980
    },
    {
      "epoch": 0.8,
      "learning_rate": 2e-05,
      "loss": 1.9928,
      "step": 985
    },
    {
      "epoch": 0.8,
      "learning_rate": 2e-05,
      "loss": 1.9927,
      "step": 990
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 2.0522,
      "step": 995
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 2.0342,
      "step": 1000
    },
    {
      "epoch": 0.81,
      "learning_rate": 2e-05,
      "loss": 2.0291,
      "step": 1005
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 2.0542,
      "step": 1010
    },
    {
      "epoch": 0.82,
      "learning_rate": 2e-05,
      "loss": 1.9575,
      "step": 1015
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 2.0268,
      "step": 1020
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 1.9652,
      "step": 1025
    },
    {
      "epoch": 0.83,
      "learning_rate": 2e-05,
      "loss": 2.0988,
      "step": 1030
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 1.9705,
      "step": 1035
    },
    {
      "epoch": 0.84,
      "learning_rate": 2e-05,
      "loss": 1.9683,
      "step": 1040
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 2.0333,
      "step": 1045
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 2.0807,
      "step": 1050
    },
    {
      "epoch": 0.85,
      "learning_rate": 2e-05,
      "loss": 2.0138,
      "step": 1055
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 1.9915,
      "step": 1060
    },
    {
      "epoch": 0.86,
      "learning_rate": 2e-05,
      "loss": 2.0396,
      "step": 1065
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 2.0208,
      "step": 1070
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 2.0008,
      "step": 1075
    },
    {
      "epoch": 0.87,
      "learning_rate": 2e-05,
      "loss": 1.9907,
      "step": 1080
    },
    {
      "epoch": 0.88,
      "learning_rate": 2e-05,
      "loss": 1.9137,
      "step": 1085
    },
    {
      "epoch": 0.88,
      "learning_rate": 2e-05,
      "loss": 2.0786,
      "step": 1090
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 1.9445,
      "step": 1095
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 2.0405,
      "step": 1100
    },
    {
      "epoch": 0.89,
      "learning_rate": 2e-05,
      "loss": 1.9961,
      "step": 1105
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 1.9778,
      "step": 1110
    },
    {
      "epoch": 0.9,
      "learning_rate": 2e-05,
      "loss": 1.9785,
      "step": 1115
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 2.0893,
      "step": 1120
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 2.0206,
      "step": 1125
    },
    {
      "epoch": 0.91,
      "learning_rate": 2e-05,
      "loss": 2.0435,
      "step": 1130
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 1.9512,
      "step": 1135
    },
    {
      "epoch": 0.92,
      "learning_rate": 2e-05,
      "loss": 2.0032,
      "step": 1140
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 1.9493,
      "step": 1145
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 1.9244,
      "step": 1150
    },
    {
      "epoch": 0.93,
      "learning_rate": 2e-05,
      "loss": 2.0021,
      "step": 1155
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 1.9534,
      "step": 1160
    },
    {
      "epoch": 0.94,
      "learning_rate": 2e-05,
      "loss": 1.925,
      "step": 1165
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 2.1044,
      "step": 1170
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 2.0376,
      "step": 1175
    },
    {
      "epoch": 0.95,
      "learning_rate": 2e-05,
      "loss": 2.0542,
      "step": 1180
    },
    {
      "epoch": 0.96,
      "learning_rate": 2e-05,
      "loss": 1.9229,
      "step": 1185
    },
    {
      "epoch": 0.96,
      "learning_rate": 2e-05,
      "loss": 1.952,
      "step": 1190
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 1.9914,
      "step": 1195
    },
    {
      "epoch": 0.97,
      "learning_rate": 2e-05,
      "loss": 1.9991,
      "step": 1200
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 1.9702,
      "step": 1205
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 1.8712,
      "step": 1210
    },
    {
      "epoch": 0.98,
      "learning_rate": 2e-05,
      "loss": 2.0369,
      "step": 1215
    },
    {
      "epoch": 0.99,
      "learning_rate": 2e-05,
      "loss": 2.1375,
      "step": 1220
    },
    {
      "epoch": 0.99,
      "learning_rate": 2e-05,
      "loss": 2.0707,
      "step": 1225
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 1.9964,
      "step": 1230
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 1.9424,
      "step": 1235
    },
    {
      "epoch": 1.0,
      "learning_rate": 2e-05,
      "loss": 1.9935,
      "step": 1240
    },
    {
      "epoch": 1.01,
      "learning_rate": 2e-05,
      "loss": 1.6404,
      "step": 1245
    },
    {
      "epoch": 1.01,
      "learning_rate": 2e-05,
      "loss": 1.5555,
      "step": 1250
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 1.5406,
      "step": 1255
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 1.6253,
      "step": 1260
    },
    {
      "epoch": 1.02,
      "learning_rate": 2e-05,
      "loss": 1.5516,
      "step": 1265
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 1.4896,
      "step": 1270
    },
    {
      "epoch": 1.03,
      "learning_rate": 2e-05,
      "loss": 1.5545,
      "step": 1275
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 1.5125,
      "step": 1280
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 1.5455,
      "step": 1285
    },
    {
      "epoch": 1.04,
      "learning_rate": 2e-05,
      "loss": 1.4915,
      "step": 1290
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 1.453,
      "step": 1295
    },
    {
      "epoch": 1.05,
      "learning_rate": 2e-05,
      "loss": 1.4764,
      "step": 1300
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 1.5592,
      "step": 1305
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 1.5472,
      "step": 1310
    },
    {
      "epoch": 1.06,
      "learning_rate": 2e-05,
      "loss": 1.4989,
      "step": 1315
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 1.4987,
      "step": 1320
    },
    {
      "epoch": 1.07,
      "learning_rate": 2e-05,
      "loss": 1.451,
      "step": 1325
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 1.5672,
      "step": 1330
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 1.5399,
      "step": 1335
    },
    {
      "epoch": 1.08,
      "learning_rate": 2e-05,
      "loss": 1.5285,
      "step": 1340
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 1.594,
      "step": 1345
    },
    {
      "epoch": 1.09,
      "learning_rate": 2e-05,
      "loss": 1.53,
      "step": 1350
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 1.5641,
      "step": 1355
    },
    {
      "epoch": 1.1,
      "learning_rate": 2e-05,
      "loss": 1.5668,
      "step": 1360
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 1.5265,
      "step": 1365
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 1.4965,
      "step": 1370
    },
    {
      "epoch": 1.11,
      "learning_rate": 2e-05,
      "loss": 1.4974,
      "step": 1375
    },
    {
      "epoch": 1.12,
      "learning_rate": 2e-05,
      "loss": 1.5232,
      "step": 1380
    },
    {
      "epoch": 1.12,
      "learning_rate": 2e-05,
      "loss": 1.5223,
      "step": 1385
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 1.5672,
      "step": 1390
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 1.5194,
      "step": 1395
    },
    {
      "epoch": 1.13,
      "learning_rate": 2e-05,
      "loss": 1.4906,
      "step": 1400
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 1.5467,
      "step": 1405
    },
    {
      "epoch": 1.14,
      "learning_rate": 2e-05,
      "loss": 1.428,
      "step": 1410
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 1.5783,
      "step": 1415
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 1.4938,
      "step": 1420
    },
    {
      "epoch": 1.15,
      "learning_rate": 2e-05,
      "loss": 1.485,
      "step": 1425
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 1.5796,
      "step": 1430
    },
    {
      "epoch": 1.16,
      "learning_rate": 2e-05,
      "loss": 1.491,
      "step": 1435
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 1.4907,
      "step": 1440
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 1.4874,
      "step": 1445
    },
    {
      "epoch": 1.17,
      "learning_rate": 2e-05,
      "loss": 1.5549,
      "step": 1450
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 1.5494,
      "step": 1455
    },
    {
      "epoch": 1.18,
      "learning_rate": 2e-05,
      "loss": 1.5035,
      "step": 1460
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 1.545,
      "step": 1465
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 1.498,
      "step": 1470
    },
    {
      "epoch": 1.19,
      "learning_rate": 2e-05,
      "loss": 1.5352,
      "step": 1475
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 1.4939,
      "step": 1480
    },
    {
      "epoch": 1.2,
      "learning_rate": 2e-05,
      "loss": 1.5376,
      "step": 1485
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 1.4745,
      "step": 1490
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 1.5165,
      "step": 1495
    },
    {
      "epoch": 1.21,
      "learning_rate": 2e-05,
      "loss": 1.4921,
      "step": 1500
    },
    {
      "epoch": 1.22,
      "learning_rate": 2e-05,
      "loss": 1.5543,
      "step": 1505
    },
    {
      "epoch": 1.22,
      "learning_rate": 2e-05,
      "loss": 1.5212,
      "step": 1510
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 1.5417,
      "step": 1515
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 1.5707,
      "step": 1520
    },
    {
      "epoch": 1.23,
      "learning_rate": 2e-05,
      "loss": 1.5293,
      "step": 1525
    },
    {
      "epoch": 1.24,
      "learning_rate": 2e-05,
      "loss": 1.5243,
      "step": 1530
    },
    {
      "epoch": 1.24,
      "learning_rate": 2e-05,
      "loss": 1.5852,
      "step": 1535
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 1.5043,
      "step": 1540
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 1.4132,
      "step": 1545
    },
    {
      "epoch": 1.25,
      "learning_rate": 2e-05,
      "loss": 1.5171,
      "step": 1550
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 1.4841,
      "step": 1555
    },
    {
      "epoch": 1.26,
      "learning_rate": 2e-05,
      "loss": 1.5396,
      "step": 1560
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 1.5959,
      "step": 1565
    },
    {
      "epoch": 1.27,
      "learning_rate": 2e-05,
      "loss": 1.5522,
      "step": 1570
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 1.5074,
      "step": 1575
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 1.5717,
      "step": 1580
    },
    {
      "epoch": 1.28,
      "learning_rate": 2e-05,
      "loss": 1.6052,
      "step": 1585
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 1.5021,
      "step": 1590
    },
    {
      "epoch": 1.29,
      "learning_rate": 2e-05,
      "loss": 1.5688,
      "step": 1595
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 1.5272,
      "step": 1600
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 1.6127,
      "step": 1605
    },
    {
      "epoch": 1.3,
      "learning_rate": 2e-05,
      "loss": 1.5163,
      "step": 1610
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 1.4874,
      "step": 1615
    },
    {
      "epoch": 1.31,
      "learning_rate": 2e-05,
      "loss": 1.4935,
      "step": 1620
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 1.5667,
      "step": 1625
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 1.5686,
      "step": 1630
    },
    {
      "epoch": 1.32,
      "learning_rate": 2e-05,
      "loss": 1.5291,
      "step": 1635
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 1.5299,
      "step": 1640
    },
    {
      "epoch": 1.33,
      "learning_rate": 2e-05,
      "loss": 1.5583,
      "step": 1645
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 1.5004,
      "step": 1650
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 1.552,
      "step": 1655
    },
    {
      "epoch": 1.34,
      "learning_rate": 2e-05,
      "loss": 1.5334,
      "step": 1660
    },
    {
      "epoch": 1.35,
      "learning_rate": 2e-05,
      "loss": 1.5249,
      "step": 1665
    },
    {
      "epoch": 1.35,
      "learning_rate": 2e-05,
      "loss": 1.5694,
      "step": 1670
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 1.4854,
      "step": 1675
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 1.4816,
      "step": 1680
    },
    {
      "epoch": 1.36,
      "learning_rate": 2e-05,
      "loss": 1.5268,
      "step": 1685
    },
    {
      "epoch": 1.37,
      "learning_rate": 2e-05,
      "loss": 1.4951,
      "step": 1690
    },
    {
      "epoch": 1.37,
      "learning_rate": 2e-05,
      "loss": 1.5026,
      "step": 1695
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 1.5675,
      "step": 1700
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 1.5503,
      "step": 1705
    },
    {
      "epoch": 1.38,
      "learning_rate": 2e-05,
      "loss": 1.6134,
      "step": 1710
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 1.5292,
      "step": 1715
    },
    {
      "epoch": 1.39,
      "learning_rate": 2e-05,
      "loss": 1.4991,
      "step": 1720
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 1.5183,
      "step": 1725
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 1.5103,
      "step": 1730
    },
    {
      "epoch": 1.4,
      "learning_rate": 2e-05,
      "loss": 1.5129,
      "step": 1735
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 1.6066,
      "step": 1740
    },
    {
      "epoch": 1.41,
      "learning_rate": 2e-05,
      "loss": 1.6065,
      "step": 1745
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 1.5733,
      "step": 1750
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 1.4913,
      "step": 1755
    },
    {
      "epoch": 1.42,
      "learning_rate": 2e-05,
      "loss": 1.5468,
      "step": 1760
    },
    {
      "epoch": 1.43,
      "learning_rate": 2e-05,
      "loss": 1.596,
      "step": 1765
    },
    {
      "epoch": 1.43,
      "learning_rate": 2e-05,
      "loss": 1.49,
      "step": 1770
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 1.5589,
      "step": 1775
    },
    {
      "epoch": 1.44,
      "learning_rate": 2e-05,
      "loss": 1.5661,
      "step": 1780
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 1.5703,
      "step": 1785
    },
    {
      "epoch": 1.45,
      "learning_rate": 2e-05,
      "loss": 1.6092,
      "step": 1790
    }
  ],
  "max_steps": 2470,
  "num_train_epochs": 2,
  "total_flos": 8.862715475793019e+17,
  "trial_name": null,
  "trial_params": null
}