{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.995790064552343,
  "eval_steps": 500,
  "global_step": 2225,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 1.5563,
      "step": 8
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.666666666666667e-06,
      "loss": 1.4423,
      "step": 16
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.3027,
      "step": 24
    },
    {
      "epoch": 0.07,
      "learning_rate": 5.333333333333334e-06,
      "loss": 1.2535,
      "step": 32
    },
    {
      "epoch": 0.09,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.2352,
      "step": 40
    },
    {
      "epoch": 0.11,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.2077,
      "step": 48
    },
    {
      "epoch": 0.13,
      "learning_rate": 9.333333333333334e-06,
      "loss": 1.1871,
      "step": 56
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 1.1708,
      "step": 64
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.2e-05,
      "loss": 1.1666,
      "step": 72
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 1.1394,
      "step": 80
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.4666666666666668e-05,
      "loss": 1.1644,
      "step": 88
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 1.1521,
      "step": 96
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 1.142,
      "step": 104
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.866666666666667e-05,
      "loss": 1.1314,
      "step": 112
    },
    {
      "epoch": 0.27,
      "learning_rate": 2e-05,
      "loss": 1.1495,
      "step": 120
    },
    {
      "epoch": 0.29,
      "learning_rate": 2.1333333333333335e-05,
      "loss": 1.1468,
      "step": 128
    },
    {
      "epoch": 0.31,
      "learning_rate": 2.2666666666666668e-05,
      "loss": 1.1457,
      "step": 136
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.4e-05,
      "loss": 1.115,
      "step": 144
    },
    {
      "epoch": 0.34,
      "learning_rate": 2.5333333333333337e-05,
      "loss": 1.1279,
      "step": 152
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 1.1145,
      "step": 160
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 1.1209,
      "step": 168
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.9333333333333336e-05,
      "loss": 1.1295,
      "step": 176
    },
    {
      "epoch": 0.41,
      "learning_rate": 3.066666666666667e-05,
      "loss": 1.1249,
      "step": 184
    },
    {
      "epoch": 0.43,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 1.1294,
      "step": 192
    },
    {
      "epoch": 0.45,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.103,
      "step": 200
    },
    {
      "epoch": 0.47,
      "learning_rate": 3.466666666666667e-05,
      "loss": 1.1241,
      "step": 208
    },
    {
      "epoch": 0.48,
      "learning_rate": 3.6e-05,
      "loss": 1.1306,
      "step": 216
    },
    {
      "epoch": 0.5,
      "learning_rate": 3.733333333333334e-05,
      "loss": 1.1096,
      "step": 224
    },
    {
      "epoch": 0.52,
      "learning_rate": 3.866666666666667e-05,
      "loss": 1.1242,
      "step": 232
    },
    {
      "epoch": 0.54,
      "learning_rate": 4e-05,
      "loss": 1.1109,
      "step": 240
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.133333333333333e-05,
      "loss": 1.1091,
      "step": 248
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.266666666666667e-05,
      "loss": 1.1304,
      "step": 256
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.1149,
      "step": 264
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.5333333333333335e-05,
      "loss": 1.106,
      "step": 272
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.666666666666667e-05,
      "loss": 1.1139,
      "step": 280
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.8e-05,
      "loss": 1.1174,
      "step": 288
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.933333333333334e-05,
      "loss": 1.0849,
      "step": 296
    },
    {
      "epoch": 0.68,
      "learning_rate": 5e-05,
      "loss": 1.1107,
      "step": 304
    },
    {
      "epoch": 0.7,
      "learning_rate": 5e-05,
      "loss": 1.0828,
      "step": 312
    },
    {
      "epoch": 0.72,
      "learning_rate": 5e-05,
      "loss": 1.1032,
      "step": 320
    },
    {
      "epoch": 0.74,
      "learning_rate": 5e-05,
      "loss": 1.15,
      "step": 328
    },
    {
      "epoch": 0.75,
      "learning_rate": 5e-05,
      "loss": 1.1088,
      "step": 336
    },
    {
      "epoch": 0.77,
      "learning_rate": 5e-05,
      "loss": 1.1008,
      "step": 344
    },
    {
      "epoch": 0.79,
      "learning_rate": 5e-05,
      "loss": 1.1174,
      "step": 352
    },
    {
      "epoch": 0.81,
      "learning_rate": 5e-05,
      "loss": 1.1043,
      "step": 360
    },
    {
      "epoch": 0.83,
      "learning_rate": 5e-05,
      "loss": 1.1196,
      "step": 368
    },
    {
      "epoch": 0.84,
      "learning_rate": 5e-05,
      "loss": 1.1339,
      "step": 376
    },
    {
      "epoch": 0.86,
      "learning_rate": 5e-05,
      "loss": 1.1186,
      "step": 384
    },
    {
      "epoch": 0.88,
      "learning_rate": 5e-05,
      "loss": 1.0837,
      "step": 392
    },
    {
      "epoch": 0.9,
      "learning_rate": 5e-05,
      "loss": 1.1105,
      "step": 400
    },
    {
      "epoch": 0.92,
      "learning_rate": 5e-05,
      "loss": 1.1249,
      "step": 408
    },
    {
      "epoch": 0.93,
      "learning_rate": 5e-05,
      "loss": 1.0846,
      "step": 416
    },
    {
      "epoch": 0.95,
      "learning_rate": 5e-05,
      "loss": 1.0925,
      "step": 424
    },
    {
      "epoch": 0.97,
      "learning_rate": 5e-05,
      "loss": 1.0917,
      "step": 432
    },
    {
      "epoch": 0.99,
      "learning_rate": 5e-05,
      "loss": 1.1177,
      "step": 440
    },
    {
      "epoch": 1.01,
      "learning_rate": 5e-05,
      "loss": 1.0964,
      "step": 448
    },
    {
      "epoch": 1.02,
      "learning_rate": 5e-05,
      "loss": 1.0494,
      "step": 456
    },
    {
      "epoch": 1.04,
      "learning_rate": 5e-05,
      "loss": 1.0444,
      "step": 464
    },
    {
      "epoch": 1.06,
      "learning_rate": 5e-05,
      "loss": 1.043,
      "step": 472
    },
    {
      "epoch": 1.08,
      "learning_rate": 5e-05,
      "loss": 1.0303,
      "step": 480
    },
    {
      "epoch": 1.1,
      "learning_rate": 5e-05,
      "loss": 1.0862,
      "step": 488
    },
    {
      "epoch": 1.11,
      "learning_rate": 5e-05,
      "loss": 1.03,
      "step": 496
    },
    {
      "epoch": 1.13,
      "learning_rate": 5e-05,
      "loss": 1.0122,
      "step": 504
    },
    {
      "epoch": 1.15,
      "learning_rate": 5e-05,
      "loss": 1.0202,
      "step": 512
    },
    {
      "epoch": 1.17,
      "learning_rate": 5e-05,
      "loss": 1.0331,
      "step": 520
    },
    {
      "epoch": 1.19,
      "learning_rate": 5e-05,
      "loss": 1.0644,
      "step": 528
    },
    {
      "epoch": 1.2,
      "learning_rate": 5e-05,
      "loss": 1.0172,
      "step": 536
    },
    {
      "epoch": 1.22,
      "learning_rate": 5e-05,
      "loss": 1.0436,
      "step": 544
    },
    {
      "epoch": 1.24,
      "learning_rate": 5e-05,
      "loss": 1.0495,
      "step": 552
    },
    {
      "epoch": 1.26,
      "learning_rate": 5e-05,
      "loss": 1.0474,
      "step": 560
    },
    {
      "epoch": 1.28,
      "learning_rate": 5e-05,
      "loss": 1.0436,
      "step": 568
    },
    {
      "epoch": 1.29,
      "learning_rate": 5e-05,
      "loss": 1.0226,
      "step": 576
    },
    {
      "epoch": 1.31,
      "learning_rate": 5e-05,
      "loss": 1.0431,
      "step": 584
    },
    {
      "epoch": 1.33,
      "learning_rate": 5e-05,
      "loss": 1.0407,
      "step": 592
    },
    {
      "epoch": 1.35,
      "learning_rate": 5e-05,
      "loss": 1.033,
      "step": 600
    },
    {
      "epoch": 1.37,
      "learning_rate": 5e-05,
      "loss": 1.0202,
      "step": 608
    },
    {
      "epoch": 1.38,
      "learning_rate": 5e-05,
      "loss": 1.0791,
      "step": 616
    },
    {
      "epoch": 1.4,
      "learning_rate": 5e-05,
      "loss": 1.0281,
      "step": 624
    },
    {
      "epoch": 1.42,
      "learning_rate": 5e-05,
      "loss": 1.0793,
      "step": 632
    },
    {
      "epoch": 1.44,
      "learning_rate": 5e-05,
      "loss": 1.0557,
      "step": 640
    },
    {
      "epoch": 1.45,
      "learning_rate": 5e-05,
      "loss": 1.0413,
      "step": 648
    },
    {
      "epoch": 1.47,
      "learning_rate": 5e-05,
      "loss": 1.0679,
      "step": 656
    },
    {
      "epoch": 1.49,
      "learning_rate": 5e-05,
      "loss": 1.0334,
      "step": 664
    },
    {
      "epoch": 1.51,
      "learning_rate": 5e-05,
      "loss": 1.0163,
      "step": 672
    },
    {
      "epoch": 1.53,
      "learning_rate": 5e-05,
      "loss": 1.0454,
      "step": 680
    },
    {
      "epoch": 1.54,
      "learning_rate": 5e-05,
      "loss": 1.0489,
      "step": 688
    },
    {
      "epoch": 1.56,
      "learning_rate": 5e-05,
      "loss": 1.0437,
      "step": 696
    },
    {
      "epoch": 1.58,
      "learning_rate": 5e-05,
      "loss": 1.0278,
      "step": 704
    },
    {
      "epoch": 1.6,
      "learning_rate": 5e-05,
      "loss": 1.0165,
      "step": 712
    },
    {
      "epoch": 1.62,
      "learning_rate": 5e-05,
      "loss": 1.0673,
      "step": 720
    },
    {
      "epoch": 1.63,
      "learning_rate": 5e-05,
      "loss": 1.0469,
      "step": 728
    },
    {
      "epoch": 1.65,
      "learning_rate": 5e-05,
      "loss": 1.0453,
      "step": 736
    },
    {
      "epoch": 1.67,
      "learning_rate": 5e-05,
      "loss": 1.0523,
      "step": 744
    },
    {
      "epoch": 1.69,
      "learning_rate": 5e-05,
      "loss": 1.0433,
      "step": 752
    },
    {
      "epoch": 1.71,
      "learning_rate": 5e-05,
      "loss": 1.0156,
      "step": 760
    },
    {
      "epoch": 1.72,
      "learning_rate": 5e-05,
      "loss": 1.0424,
      "step": 768
    },
    {
      "epoch": 1.74,
      "learning_rate": 5e-05,
      "loss": 1.0216,
      "step": 776
    },
    {
      "epoch": 1.76,
      "learning_rate": 5e-05,
      "loss": 1.0369,
      "step": 784
    },
    {
      "epoch": 1.78,
      "learning_rate": 5e-05,
      "loss": 1.0509,
      "step": 792
    },
    {
      "epoch": 1.8,
      "learning_rate": 5e-05,
      "loss": 1.0619,
      "step": 800
    },
    {
      "epoch": 1.81,
      "learning_rate": 5e-05,
      "loss": 1.0164,
      "step": 808
    },
    {
      "epoch": 1.83,
      "learning_rate": 5e-05,
      "loss": 1.0256,
      "step": 816
    },
    {
      "epoch": 1.85,
      "learning_rate": 5e-05,
      "loss": 1.0028,
      "step": 824
    },
    {
      "epoch": 1.87,
      "learning_rate": 5e-05,
      "loss": 1.0244,
      "step": 832
    },
    {
      "epoch": 1.89,
      "learning_rate": 5e-05,
      "loss": 1.0494,
      "step": 840
    },
    {
      "epoch": 1.9,
      "learning_rate": 5e-05,
      "loss": 1.0387,
      "step": 848
    },
    {
      "epoch": 1.92,
      "learning_rate": 5e-05,
      "loss": 1.0168,
      "step": 856
    },
    {
      "epoch": 1.94,
      "learning_rate": 5e-05,
      "loss": 1.0455,
      "step": 864
    },
    {
      "epoch": 1.96,
      "learning_rate": 5e-05,
      "loss": 1.0622,
      "step": 872
    },
    {
      "epoch": 1.98,
      "learning_rate": 5e-05,
      "loss": 1.0564,
      "step": 880
    },
    {
      "epoch": 1.99,
      "learning_rate": 5e-05,
      "loss": 1.0656,
      "step": 888
    },
    {
      "epoch": 2.01,
      "learning_rate": 5e-05,
      "loss": 0.9658,
      "step": 896
    },
    {
      "epoch": 2.03,
      "learning_rate": 5e-05,
      "loss": 0.9094,
      "step": 904
    },
    {
      "epoch": 2.05,
      "learning_rate": 5e-05,
      "loss": 0.9343,
      "step": 912
    },
    {
      "epoch": 2.07,
      "learning_rate": 5e-05,
      "loss": 0.8941,
      "step": 920
    },
    {
      "epoch": 2.08,
      "learning_rate": 5e-05,
      "loss": 0.8903,
      "step": 928
    },
    {
      "epoch": 2.1,
      "learning_rate": 5e-05,
      "loss": 0.892,
      "step": 936
    },
    {
      "epoch": 2.12,
      "learning_rate": 5e-05,
      "loss": 0.9425,
      "step": 944
    },
    {
      "epoch": 2.14,
      "learning_rate": 5e-05,
      "loss": 0.932,
      "step": 952
    },
    {
      "epoch": 2.16,
      "learning_rate": 5e-05,
      "loss": 0.9171,
      "step": 960
    },
    {
      "epoch": 2.17,
      "learning_rate": 5e-05,
      "loss": 0.8976,
      "step": 968
    },
    {
      "epoch": 2.19,
      "learning_rate": 5e-05,
      "loss": 0.9055,
      "step": 976
    },
    {
      "epoch": 2.21,
      "learning_rate": 5e-05,
      "loss": 0.9442,
      "step": 984
    },
    {
      "epoch": 2.23,
      "learning_rate": 5e-05,
      "loss": 0.8948,
      "step": 992
    },
    {
      "epoch": 2.25,
      "learning_rate": 5e-05,
      "loss": 0.9297,
      "step": 1000
    },
    {
      "epoch": 2.26,
      "learning_rate": 5e-05,
      "loss": 0.8957,
      "step": 1008
    },
    {
      "epoch": 2.28,
      "learning_rate": 5e-05,
      "loss": 0.9264,
      "step": 1016
    },
    {
      "epoch": 2.3,
      "learning_rate": 5e-05,
      "loss": 0.9391,
      "step": 1024
    },
    {
      "epoch": 2.32,
      "learning_rate": 5e-05,
      "loss": 0.8936,
      "step": 1032
    },
    {
      "epoch": 2.34,
      "learning_rate": 5e-05,
      "loss": 0.9537,
      "step": 1040
    },
    {
      "epoch": 2.35,
      "learning_rate": 5e-05,
      "loss": 0.9168,
      "step": 1048
    },
    {
      "epoch": 2.37,
      "learning_rate": 5e-05,
      "loss": 0.9092,
      "step": 1056
    },
    {
      "epoch": 2.39,
      "learning_rate": 5e-05,
      "loss": 0.9198,
      "step": 1064
    },
    {
      "epoch": 2.41,
      "learning_rate": 5e-05,
      "loss": 0.9284,
      "step": 1072
    },
    {
      "epoch": 2.42,
      "learning_rate": 5e-05,
      "loss": 0.8926,
      "step": 1080
    },
    {
      "epoch": 2.44,
      "learning_rate": 5e-05,
      "loss": 0.9338,
      "step": 1088
    },
    {
      "epoch": 2.46,
      "learning_rate": 5e-05,
      "loss": 0.91,
      "step": 1096
    },
    {
      "epoch": 2.48,
      "learning_rate": 5e-05,
      "loss": 0.9444,
      "step": 1104
    },
    {
      "epoch": 2.5,
      "learning_rate": 5e-05,
      "loss": 0.9326,
      "step": 1112
    },
    {
      "epoch": 2.51,
      "learning_rate": 5e-05,
      "loss": 0.901,
      "step": 1120
    },
    {
      "epoch": 2.53,
      "learning_rate": 5e-05,
      "loss": 0.9344,
      "step": 1128
    },
    {
      "epoch": 2.55,
      "learning_rate": 5e-05,
      "loss": 0.9344,
      "step": 1136
    },
    {
      "epoch": 2.57,
      "learning_rate": 5e-05,
      "loss": 0.9063,
      "step": 1144
    },
    {
      "epoch": 2.59,
      "learning_rate": 5e-05,
      "loss": 0.9123,
      "step": 1152
    },
    {
      "epoch": 2.6,
      "learning_rate": 5e-05,
      "loss": 0.9287,
      "step": 1160
    },
    {
      "epoch": 2.62,
      "learning_rate": 5e-05,
      "loss": 0.8998,
      "step": 1168
    },
    {
      "epoch": 2.64,
      "learning_rate": 5e-05,
      "loss": 0.9101,
      "step": 1176
    },
    {
      "epoch": 2.66,
      "learning_rate": 5e-05,
      "loss": 0.9353,
      "step": 1184
    },
    {
      "epoch": 2.68,
      "learning_rate": 5e-05,
      "loss": 0.942,
      "step": 1192
    },
    {
      "epoch": 2.69,
      "learning_rate": 5e-05,
      "loss": 0.9194,
      "step": 1200
    },
    {
      "epoch": 2.71,
      "learning_rate": 5e-05,
      "loss": 0.9221,
      "step": 1208
    },
    {
      "epoch": 2.73,
      "learning_rate": 5e-05,
      "loss": 0.9227,
      "step": 1216
    },
    {
      "epoch": 2.75,
      "learning_rate": 5e-05,
      "loss": 0.9409,
      "step": 1224
    },
    {
      "epoch": 2.77,
      "learning_rate": 5e-05,
      "loss": 0.9192,
      "step": 1232
    },
    {
      "epoch": 2.78,
      "learning_rate": 5e-05,
      "loss": 0.9274,
      "step": 1240
    },
    {
      "epoch": 2.8,
      "learning_rate": 5e-05,
      "loss": 0.9422,
      "step": 1248
    },
    {
      "epoch": 2.82,
      "learning_rate": 5e-05,
      "loss": 0.9236,
      "step": 1256
    },
    {
      "epoch": 2.84,
      "learning_rate": 5e-05,
      "loss": 0.9185,
      "step": 1264
    },
    {
      "epoch": 2.86,
      "learning_rate": 5e-05,
      "loss": 0.9179,
      "step": 1272
    },
    {
      "epoch": 2.87,
      "learning_rate": 5e-05,
      "loss": 0.9186,
      "step": 1280
    },
    {
      "epoch": 2.89,
      "learning_rate": 5e-05,
      "loss": 0.9117,
      "step": 1288
    },
    {
      "epoch": 2.91,
      "learning_rate": 5e-05,
      "loss": 0.9058,
      "step": 1296
    },
    {
      "epoch": 2.93,
      "learning_rate": 5e-05,
      "loss": 0.9486,
      "step": 1304
    },
    {
      "epoch": 2.95,
      "learning_rate": 5e-05,
      "loss": 0.9157,
      "step": 1312
    },
    {
      "epoch": 2.96,
      "learning_rate": 5e-05,
      "loss": 0.9119,
      "step": 1320
    },
    {
      "epoch": 2.98,
      "learning_rate": 5e-05,
      "loss": 0.9238,
      "step": 1328
    },
    {
      "epoch": 3.0,
      "learning_rate": 5e-05,
      "loss": 0.9201,
      "step": 1336
    },
    {
      "epoch": 3.02,
      "learning_rate": 5e-05,
      "loss": 0.7697,
      "step": 1344
    },
    {
      "epoch": 3.04,
      "learning_rate": 5e-05,
      "loss": 0.7608,
      "step": 1352
    },
    {
      "epoch": 3.05,
      "learning_rate": 5e-05,
      "loss": 0.7581,
      "step": 1360
    },
    {
      "epoch": 3.07,
      "learning_rate": 5e-05,
      "loss": 0.7518,
      "step": 1368
    },
    {
      "epoch": 3.09,
      "learning_rate": 5e-05,
      "loss": 0.7121,
      "step": 1376
    },
    {
      "epoch": 3.11,
      "learning_rate": 5e-05,
      "loss": 0.7284,
      "step": 1384
    },
    {
      "epoch": 3.13,
      "learning_rate": 5e-05,
      "loss": 0.7513,
      "step": 1392
    },
    {
      "epoch": 3.14,
      "learning_rate": 5e-05,
      "loss": 0.7529,
      "step": 1400
    },
    {
      "epoch": 3.16,
      "learning_rate": 5e-05,
      "loss": 0.7443,
      "step": 1408
    },
    {
      "epoch": 3.18,
      "learning_rate": 5e-05,
      "loss": 0.7603,
      "step": 1416
    },
    {
      "epoch": 3.2,
      "learning_rate": 5e-05,
      "loss": 0.7242,
      "step": 1424
    },
    {
      "epoch": 3.22,
      "learning_rate": 5e-05,
      "loss": 0.7504,
      "step": 1432
    },
    {
      "epoch": 3.23,
      "learning_rate": 5e-05,
      "loss": 0.7363,
      "step": 1440
    },
    {
      "epoch": 3.25,
      "learning_rate": 5e-05,
      "loss": 0.7307,
      "step": 1448
    },
    {
      "epoch": 3.27,
      "learning_rate": 5e-05,
      "loss": 0.757,
      "step": 1456
    },
    {
      "epoch": 3.29,
      "learning_rate": 5e-05,
      "loss": 0.7337,
      "step": 1464
    },
    {
      "epoch": 3.31,
      "learning_rate": 5e-05,
      "loss": 0.7338,
      "step": 1472
    },
    {
      "epoch": 3.32,
      "learning_rate": 5e-05,
      "loss": 0.7632,
      "step": 1480
    },
    {
      "epoch": 3.34,
      "learning_rate": 5e-05,
      "loss": 0.7643,
      "step": 1488
    },
    {
      "epoch": 3.36,
      "learning_rate": 5e-05,
      "loss": 0.7458,
      "step": 1496
    },
    {
      "epoch": 3.38,
      "learning_rate": 5e-05,
      "loss": 0.7704,
      "step": 1504
    },
    {
      "epoch": 3.39,
      "learning_rate": 5e-05,
      "loss": 0.755,
      "step": 1512
    },
    {
      "epoch": 3.41,
      "learning_rate": 5e-05,
      "loss": 0.7182,
      "step": 1520
    },
    {
      "epoch": 3.43,
      "learning_rate": 5e-05,
      "loss": 0.7533,
      "step": 1528
    },
    {
      "epoch": 3.45,
      "learning_rate": 5e-05,
      "loss": 0.7475,
      "step": 1536
    },
    {
      "epoch": 3.47,
      "learning_rate": 5e-05,
      "loss": 0.7671,
      "step": 1544
    },
    {
      "epoch": 3.48,
      "learning_rate": 5e-05,
      "loss": 0.7541,
      "step": 1552
    },
    {
      "epoch": 3.5,
      "learning_rate": 5e-05,
      "loss": 0.7758,
      "step": 1560
    },
    {
      "epoch": 3.52,
      "learning_rate": 5e-05,
      "loss": 0.7387,
      "step": 1568
    },
    {
      "epoch": 3.54,
      "learning_rate": 5e-05,
      "loss": 0.7928,
      "step": 1576
    },
    {
      "epoch": 3.56,
      "learning_rate": 5e-05,
      "loss": 0.7494,
      "step": 1584
    },
    {
      "epoch": 3.57,
      "learning_rate": 5e-05,
      "loss": 0.7716,
      "step": 1592
    },
    {
      "epoch": 3.59,
      "learning_rate": 5e-05,
      "loss": 0.7735,
      "step": 1600
    },
    {
      "epoch": 3.61,
      "learning_rate": 5e-05,
      "loss": 0.7688,
      "step": 1608
    },
    {
      "epoch": 3.63,
      "learning_rate": 5e-05,
      "loss": 0.7852,
      "step": 1616
    },
    {
      "epoch": 3.65,
      "learning_rate": 5e-05,
      "loss": 0.7501,
      "step": 1624
    },
    {
      "epoch": 3.66,
      "learning_rate": 5e-05,
      "loss": 0.7754,
      "step": 1632
    },
    {
      "epoch": 3.68,
      "learning_rate": 5e-05,
      "loss": 0.756,
      "step": 1640
    },
    {
      "epoch": 3.7,
      "learning_rate": 5e-05,
      "loss": 0.7778,
      "step": 1648
    },
    {
      "epoch": 3.72,
      "learning_rate": 5e-05,
      "loss": 0.7835,
      "step": 1656
    },
    {
      "epoch": 3.74,
      "learning_rate": 5e-05,
      "loss": 0.7714,
      "step": 1664
    },
    {
      "epoch": 3.75,
      "learning_rate": 5e-05,
      "loss": 0.7417,
      "step": 1672
    },
    {
      "epoch": 3.77,
      "learning_rate": 5e-05,
      "loss": 0.7803,
      "step": 1680
    },
    {
      "epoch": 3.79,
      "learning_rate": 5e-05,
      "loss": 0.7616,
      "step": 1688
    },
    {
      "epoch": 3.81,
      "learning_rate": 5e-05,
      "loss": 0.7574,
      "step": 1696
    },
    {
      "epoch": 3.83,
      "learning_rate": 5e-05,
      "loss": 0.7895,
      "step": 1704
    },
    {
      "epoch": 3.84,
      "learning_rate": 5e-05,
      "loss": 0.7773,
      "step": 1712
    },
    {
      "epoch": 3.86,
      "learning_rate": 5e-05,
      "loss": 0.7839,
      "step": 1720
    },
    {
      "epoch": 3.88,
      "learning_rate": 5e-05,
      "loss": 0.7675,
      "step": 1728
    },
    {
      "epoch": 3.9,
      "learning_rate": 5e-05,
      "loss": 0.7649,
      "step": 1736
    },
    {
      "epoch": 3.92,
      "learning_rate": 5e-05,
      "loss": 0.7656,
      "step": 1744
    },
    {
      "epoch": 3.93,
      "learning_rate": 5e-05,
      "loss": 0.7725,
      "step": 1752
    },
    {
      "epoch": 3.95,
      "learning_rate": 5e-05,
      "loss": 0.7609,
      "step": 1760
    },
    {
      "epoch": 3.97,
      "learning_rate": 5e-05,
      "loss": 0.7664,
      "step": 1768
    },
    {
      "epoch": 3.99,
      "learning_rate": 5e-05,
      "loss": 0.7945,
      "step": 1776
    },
    {
      "epoch": 4.01,
      "learning_rate": 5e-05,
      "loss": 0.7065,
      "step": 1784
    },
    {
      "epoch": 4.02,
      "learning_rate": 5e-05,
      "loss": 0.5463,
      "step": 1792
    },
    {
      "epoch": 4.04,
      "learning_rate": 5e-05,
      "loss": 0.549,
      "step": 1800
    },
    {
      "epoch": 4.06,
      "learning_rate": 5e-05,
      "loss": 0.571,
      "step": 1808
    },
    {
      "epoch": 4.08,
      "learning_rate": 5e-05,
      "loss": 0.5538,
      "step": 1816
    },
    {
      "epoch": 4.1,
      "learning_rate": 5e-05,
      "loss": 0.5649,
      "step": 1824
    },
    {
      "epoch": 4.11,
      "learning_rate": 5e-05,
      "loss": 0.5719,
      "step": 1832
    },
    {
      "epoch": 4.13,
      "learning_rate": 5e-05,
      "loss": 0.5273,
      "step": 1840
    },
    {
      "epoch": 4.15,
      "learning_rate": 5e-05,
      "loss": 0.5482,
      "step": 1848
    },
    {
      "epoch": 4.17,
      "learning_rate": 5e-05,
      "loss": 0.5708,
      "step": 1856
    },
    {
      "epoch": 4.19,
      "learning_rate": 5e-05,
      "loss": 0.5643,
      "step": 1864
    },
    {
      "epoch": 4.2,
      "learning_rate": 5e-05,
      "loss": 0.5372,
      "step": 1872
    },
    {
      "epoch": 4.22,
      "learning_rate": 5e-05,
      "loss": 0.5623,
      "step": 1880
    },
    {
      "epoch": 4.24,
      "learning_rate": 5e-05,
      "loss": 0.5869,
      "step": 1888
    },
    {
      "epoch": 4.26,
      "learning_rate": 5e-05,
      "loss": 0.5929,
      "step": 1896
    },
    {
      "epoch": 4.28,
      "learning_rate": 5e-05,
      "loss": 0.5645,
      "step": 1904
    },
    {
      "epoch": 4.29,
      "learning_rate": 5e-05,
      "loss": 0.5901,
      "step": 1912
    },
    {
      "epoch": 4.31,
      "learning_rate": 5e-05,
      "loss": 0.5585,
      "step": 1920
    },
    {
      "epoch": 4.33,
      "learning_rate": 5e-05,
      "loss": 0.571,
      "step": 1928
    },
    {
      "epoch": 4.35,
      "learning_rate": 5e-05,
      "loss": 0.5824,
      "step": 1936
    },
    {
      "epoch": 4.36,
      "learning_rate": 5e-05,
      "loss": 0.5801,
      "step": 1944
    },
    {
      "epoch": 4.38,
      "learning_rate": 5e-05,
      "loss": 0.5736,
      "step": 1952
    },
    {
      "epoch": 4.4,
      "learning_rate": 5e-05,
      "loss": 0.5874,
      "step": 1960
    },
    {
      "epoch": 4.42,
      "learning_rate": 5e-05,
      "loss": 0.5795,
      "step": 1968
    },
    {
      "epoch": 4.44,
      "learning_rate": 5e-05,
      "loss": 0.5868,
      "step": 1976
    },
    {
      "epoch": 4.45,
      "learning_rate": 5e-05,
      "loss": 0.5649,
      "step": 1984
    },
    {
      "epoch": 4.47,
      "learning_rate": 5e-05,
      "loss": 0.5769,
      "step": 1992
    },
    {
      "epoch": 4.49,
      "learning_rate": 5e-05,
      "loss": 0.5944,
      "step": 2000
    },
    {
      "epoch": 4.51,
      "learning_rate": 5e-05,
      "loss": 0.5876,
      "step": 2008
    },
    {
      "epoch": 4.53,
      "learning_rate": 5e-05,
      "loss": 0.6052,
      "step": 2016
    },
    {
      "epoch": 4.54,
      "learning_rate": 5e-05,
      "loss": 0.5994,
      "step": 2024
    },
    {
      "epoch": 4.56,
      "learning_rate": 5e-05,
      "loss": 0.5684,
      "step": 2032
    },
    {
      "epoch": 4.58,
      "learning_rate": 5e-05,
      "loss": 0.5927,
      "step": 2040
    },
    {
      "epoch": 4.6,
      "learning_rate": 5e-05,
      "loss": 0.5857,
      "step": 2048
    },
    {
      "epoch": 4.62,
      "learning_rate": 5e-05,
      "loss": 0.566,
      "step": 2056
    },
    {
      "epoch": 4.63,
      "learning_rate": 5e-05,
      "loss": 0.5763,
      "step": 2064
    },
    {
      "epoch": 4.65,
      "learning_rate": 5e-05,
      "loss": 0.5997,
      "step": 2072
    },
    {
      "epoch": 4.67,
      "learning_rate": 5e-05,
      "loss": 0.5655,
      "step": 2080
    },
    {
      "epoch": 4.69,
      "learning_rate": 5e-05,
      "loss": 0.5859,
      "step": 2088
    },
    {
      "epoch": 4.71,
      "learning_rate": 5e-05,
      "loss": 0.5636,
      "step": 2096
    },
    {
      "epoch": 4.72,
      "learning_rate": 5e-05,
      "loss": 0.5964,
      "step": 2104
    },
    {
      "epoch": 4.74,
      "learning_rate": 5e-05,
      "loss": 0.5534,
      "step": 2112
    },
    {
      "epoch": 4.76,
      "learning_rate": 5e-05,
      "loss": 0.5858,
      "step": 2120
    },
    {
      "epoch": 4.78,
      "learning_rate": 5e-05,
      "loss": 0.5916,
      "step": 2128
    },
    {
      "epoch": 4.8,
      "learning_rate": 5e-05,
      "loss": 0.5862,
      "step": 2136
    },
    {
      "epoch": 4.81,
      "learning_rate": 5e-05,
      "loss": 0.6118,
      "step": 2144
    },
    {
      "epoch": 4.83,
      "learning_rate": 5e-05,
      "loss": 0.6023,
      "step": 2152
    },
    {
      "epoch": 4.85,
      "learning_rate": 5e-05,
      "loss": 0.5927,
      "step": 2160
    },
    {
      "epoch": 4.87,
      "learning_rate": 5e-05,
      "loss": 0.5911,
      "step": 2168
    },
    {
      "epoch": 4.89,
      "learning_rate": 5e-05,
      "loss": 0.6043,
      "step": 2176
    },
    {
      "epoch": 4.9,
      "learning_rate": 5e-05,
      "loss": 0.5779,
      "step": 2184
    },
    {
      "epoch": 4.92,
      "learning_rate": 5e-05,
      "loss": 0.6112,
      "step": 2192
    },
    {
      "epoch": 4.94,
      "learning_rate": 5e-05,
      "loss": 0.5893,
      "step": 2200
    },
    {
      "epoch": 4.96,
      "learning_rate": 5e-05,
      "loss": 0.597,
      "step": 2208
    },
    {
      "epoch": 4.98,
      "learning_rate": 5e-05,
      "loss": 0.5992,
      "step": 2216
    },
    {
      "epoch": 4.99,
      "learning_rate": 5e-05,
      "loss": 0.5933,
      "step": 2224
    }
  ],
  "logging_steps": 8,
  "max_steps": 2225,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 1.5950431556342907e+18,
  "trial_name": null,
  "trial_params": null
}