{
  "best_metric": 0.24075740575790405,
  "best_model_checkpoint": "./results/checkpoint-6750",
  "epoch": 3.0,
  "global_step": 6750,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 4.9992592592592596e-05,
      "loss": 6.945,
      "step": 1
    },
    {
      "epoch": 0.0,
      "learning_rate": 4.9925925925925926e-05,
      "loss": 1.4118,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.9851851851851855e-05,
      "loss": 0.7118,
      "step": 20
    },
    {
      "epoch": 0.01,
      "eval_loss": 0.6554884910583496,
      "eval_runtime": 6.9152,
      "eval_samples_per_second": 144.754,
      "eval_steps_per_second": 36.297,
      "step": 25
    },
    {
      "epoch": 0.01,
      "learning_rate": 4.977777777777778e-05,
      "loss": 0.8383,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.970370370370371e-05,
      "loss": 0.6623,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 4.962962962962963e-05,
      "loss": 0.6282,
      "step": 50
    },
    {
      "epoch": 0.02,
      "eval_loss": 0.5736271739006042,
      "eval_runtime": 6.8557,
      "eval_samples_per_second": 146.01,
      "eval_steps_per_second": 36.612,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.955555555555556e-05,
      "loss": 0.5446,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 4.9481481481481485e-05,
      "loss": 0.5983,
      "step": 70
    },
    {
      "epoch": 0.03,
      "eval_loss": 0.5498507022857666,
      "eval_runtime": 6.8866,
      "eval_samples_per_second": 145.356,
      "eval_steps_per_second": 36.448,
      "step": 75
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.940740740740741e-05,
      "loss": 0.5406,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.933333333333334e-05,
      "loss": 0.618,
      "step": 90
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.925925925925926e-05,
      "loss": 0.6346,
      "step": 100
    },
    {
      "epoch": 0.04,
      "eval_loss": 0.5268338918685913,
      "eval_runtime": 6.9999,
      "eval_samples_per_second": 143.001,
      "eval_steps_per_second": 35.857,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.918518518518519e-05,
      "loss": 0.53,
      "step": 110
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.9111111111111114e-05,
      "loss": 0.4268,
      "step": 120
    },
    {
      "epoch": 0.06,
      "eval_loss": 0.5140772461891174,
      "eval_runtime": 7.002,
      "eval_samples_per_second": 142.959,
      "eval_steps_per_second": 35.847,
      "step": 125
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.903703703703704e-05,
      "loss": 0.5474,
      "step": 130
    },
    {
      "epoch": 0.06,
      "learning_rate": 4.896296296296297e-05,
      "loss": 0.5773,
      "step": 140
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.888888888888889e-05,
      "loss": 0.5611,
      "step": 150
    },
    {
      "epoch": 0.07,
      "eval_loss": 0.5012452006340027,
      "eval_runtime": 7.2224,
      "eval_samples_per_second": 138.597,
      "eval_steps_per_second": 34.753,
      "step": 150
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.881481481481482e-05,
      "loss": 0.5421,
      "step": 160
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.874074074074074e-05,
      "loss": 0.5056,
      "step": 170
    },
    {
      "epoch": 0.08,
      "eval_loss": 0.492546945810318,
      "eval_runtime": 7.468,
      "eval_samples_per_second": 134.038,
      "eval_steps_per_second": 33.61,
      "step": 175
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.866666666666667e-05,
      "loss": 0.5086,
      "step": 180
    },
    {
      "epoch": 0.08,
      "learning_rate": 4.8592592592592596e-05,
      "loss": 0.4423,
      "step": 190
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.851851851851852e-05,
      "loss": 0.5391,
      "step": 200
    },
    {
      "epoch": 0.09,
      "eval_loss": 0.4847224950790405,
      "eval_runtime": 7.3642,
      "eval_samples_per_second": 135.928,
      "eval_steps_per_second": 34.084,
      "step": 200
    },
    {
      "epoch": 0.09,
      "learning_rate": 4.844444444444445e-05,
      "loss": 0.4785,
      "step": 210
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.837037037037037e-05,
      "loss": 0.4303,
      "step": 220
    },
    {
      "epoch": 0.1,
      "eval_loss": 0.4781314730644226,
      "eval_runtime": 7.4029,
      "eval_samples_per_second": 135.217,
      "eval_steps_per_second": 33.906,
      "step": 225
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.82962962962963e-05,
      "loss": 0.5914,
      "step": 230
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.8222222222222225e-05,
      "loss": 0.4892,
      "step": 240
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.814814814814815e-05,
      "loss": 0.5442,
      "step": 250
    },
    {
      "epoch": 0.11,
      "eval_loss": 0.4688411355018616,
      "eval_runtime": 7.5871,
      "eval_samples_per_second": 131.935,
      "eval_steps_per_second": 33.083,
      "step": 250
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.807407407407408e-05,
      "loss": 0.4823,
      "step": 260
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.8e-05,
      "loss": 0.4739,
      "step": 270
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.462319940328598,
      "eval_runtime": 7.2477,
      "eval_samples_per_second": 138.112,
      "eval_steps_per_second": 34.632,
      "step": 275
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.792592592592593e-05,
      "loss": 0.5542,
      "step": 280
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.7851851851851854e-05,
      "loss": 0.4593,
      "step": 290
    },
    {
      "epoch": 0.13,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 0.4388,
      "step": 300
    },
    {
      "epoch": 0.13,
      "eval_loss": 0.4547964036464691,
      "eval_runtime": 7.2941,
      "eval_samples_per_second": 137.235,
      "eval_steps_per_second": 34.412,
      "step": 300
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.770370370370371e-05,
      "loss": 0.4762,
      "step": 310
    },
    {
      "epoch": 0.14,
      "learning_rate": 4.762962962962963e-05,
      "loss": 0.428,
      "step": 320
    },
    {
      "epoch": 0.14,
      "eval_loss": 0.45131900906562805,
      "eval_runtime": 7.2669,
      "eval_samples_per_second": 137.748,
      "eval_steps_per_second": 34.54,
      "step": 325
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.755555555555556e-05,
      "loss": 0.4719,
      "step": 330
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.7481481481481483e-05,
      "loss": 0.4293,
      "step": 340
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.740740740740741e-05,
      "loss": 0.4112,
      "step": 350
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.4421839714050293,
      "eval_runtime": 7.3019,
      "eval_samples_per_second": 137.087,
      "eval_steps_per_second": 34.374,
      "step": 350
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7333333333333336e-05,
      "loss": 0.3981,
      "step": 360
    },
    {
      "epoch": 0.16,
      "learning_rate": 4.7259259259259266e-05,
      "loss": 0.501,
      "step": 370
    },
    {
      "epoch": 0.17,
      "eval_loss": 0.43612053990364075,
      "eval_runtime": 7.1757,
      "eval_samples_per_second": 139.499,
      "eval_steps_per_second": 34.979,
      "step": 375
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.718518518518519e-05,
      "loss": 0.4066,
      "step": 380
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.711111111111111e-05,
      "loss": 0.4892,
      "step": 390
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.703703703703704e-05,
      "loss": 0.4584,
      "step": 400
    },
    {
      "epoch": 0.18,
      "eval_loss": 0.4388324022293091,
      "eval_runtime": 7.2341,
      "eval_samples_per_second": 138.372,
      "eval_steps_per_second": 34.697,
      "step": 400
    },
    {
      "epoch": 0.18,
      "learning_rate": 4.6962962962962966e-05,
      "loss": 0.432,
      "step": 410
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.6888888888888895e-05,
      "loss": 0.4717,
      "step": 420
    },
    {
      "epoch": 0.19,
      "eval_loss": 0.42829033732414246,
      "eval_runtime": 7.261,
      "eval_samples_per_second": 137.86,
      "eval_steps_per_second": 34.568,
      "step": 425
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.681481481481482e-05,
      "loss": 0.399,
      "step": 430
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.674074074074074e-05,
      "loss": 0.4021,
      "step": 440
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.4122,
      "step": 450
    },
    {
      "epoch": 0.2,
      "eval_loss": 0.4233491122722626,
      "eval_runtime": 7.2994,
      "eval_samples_per_second": 137.134,
      "eval_steps_per_second": 34.386,
      "step": 450
    },
    {
      "epoch": 0.2,
      "learning_rate": 4.6592592592592595e-05,
      "loss": 0.4539,
      "step": 460
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.6518518518518525e-05,
      "loss": 0.5041,
      "step": 470
    },
    {
      "epoch": 0.21,
      "eval_loss": 0.4170687794685364,
      "eval_runtime": 7.1967,
      "eval_samples_per_second": 139.092,
      "eval_steps_per_second": 34.877,
      "step": 475
    },
    {
      "epoch": 0.21,
      "learning_rate": 4.644444444444445e-05,
      "loss": 0.5674,
      "step": 480
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.637037037037038e-05,
      "loss": 0.4226,
      "step": 490
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.62962962962963e-05,
      "loss": 0.4153,
      "step": 500
    },
    {
      "epoch": 0.22,
      "eval_loss": 0.41686439514160156,
      "eval_runtime": 7.2526,
      "eval_samples_per_second": 138.019,
      "eval_steps_per_second": 34.608,
      "step": 500
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.6222222222222224e-05,
      "loss": 0.3956,
      "step": 510
    },
    {
      "epoch": 0.23,
      "learning_rate": 4.6148148148148154e-05,
      "loss": 0.4524,
      "step": 520
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.41372087597846985,
      "eval_runtime": 7.2599,
      "eval_samples_per_second": 137.88,
      "eval_steps_per_second": 34.573,
      "step": 525
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.607407407407408e-05,
      "loss": 0.4319,
      "step": 530
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4098,
      "step": 540
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.592592592592593e-05,
      "loss": 0.511,
      "step": 550
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.406745046377182,
      "eval_runtime": 7.4218,
      "eval_samples_per_second": 134.872,
      "eval_steps_per_second": 33.819,
      "step": 550
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.585185185185185e-05,
      "loss": 0.5086,
      "step": 560
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.577777777777778e-05,
      "loss": 0.4096,
      "step": 570
    },
    {
      "epoch": 0.26,
      "eval_loss": 0.404786080121994,
      "eval_runtime": 7.39,
      "eval_samples_per_second": 135.453,
      "eval_steps_per_second": 33.965,
      "step": 575
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.5703703703703706e-05,
      "loss": 0.4281,
      "step": 580
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.5629629629629636e-05,
      "loss": 0.4779,
      "step": 590
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.555555555555556e-05,
      "loss": 0.405,
      "step": 600
    },
    {
      "epoch": 0.27,
      "eval_loss": 0.40296873450279236,
      "eval_runtime": 7.4985,
      "eval_samples_per_second": 133.493,
      "eval_steps_per_second": 33.473,
      "step": 600
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.548148148148149e-05,
      "loss": 0.4144,
      "step": 610
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.540740740740741e-05,
      "loss": 0.3645,
      "step": 620
    },
    {
      "epoch": 0.28,
      "eval_loss": 0.39786386489868164,
      "eval_runtime": 7.5248,
      "eval_samples_per_second": 133.027,
      "eval_steps_per_second": 33.356,
      "step": 625
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.5333333333333335e-05,
      "loss": 0.3679,
      "step": 630
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.5259259259259265e-05,
      "loss": 0.3724,
      "step": 640
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.518518518518519e-05,
      "loss": 0.4452,
      "step": 650
    },
    {
      "epoch": 0.29,
      "eval_loss": 0.39415648579597473,
      "eval_runtime": 7.2333,
      "eval_samples_per_second": 138.388,
      "eval_steps_per_second": 34.701,
      "step": 650
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.511111111111112e-05,
      "loss": 0.325,
      "step": 660
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.503703703703704e-05,
      "loss": 0.4001,
      "step": 670
    },
    {
      "epoch": 0.3,
      "eval_loss": 0.38874566555023193,
      "eval_runtime": 7.2318,
      "eval_samples_per_second": 138.417,
      "eval_steps_per_second": 34.708,
      "step": 675
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.496296296296297e-05,
      "loss": 0.3631,
      "step": 680
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.4888888888888894e-05,
      "loss": 0.4206,
      "step": 690
    },
    {
      "epoch": 0.31,
      "learning_rate": 4.481481481481482e-05,
      "loss": 0.4616,
      "step": 700
    },
    {
      "epoch": 0.31,
      "eval_loss": 0.38908329606056213,
      "eval_runtime": 7.2894,
      "eval_samples_per_second": 137.323,
      "eval_steps_per_second": 34.434,
      "step": 700
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.474074074074075e-05,
      "loss": 0.4723,
      "step": 710
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.466666666666667e-05,
      "loss": 0.3344,
      "step": 720
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.38453349471092224,
      "eval_runtime": 7.2247,
      "eval_samples_per_second": 138.552,
      "eval_steps_per_second": 34.742,
      "step": 725
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.4592592592592594e-05,
      "loss": 0.4402,
      "step": 730
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.4518518518518523e-05,
      "loss": 0.389,
      "step": 740
    },
    {
      "epoch": 0.33,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.3899,
      "step": 750
    },
    {
      "epoch": 0.33,
      "eval_loss": 0.3842224180698395,
      "eval_runtime": 7.3669,
      "eval_samples_per_second": 135.879,
      "eval_steps_per_second": 34.071,
      "step": 750
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.4370370370370376e-05,
      "loss": 0.3759,
      "step": 760
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.42962962962963e-05,
      "loss": 0.3662,
      "step": 770
    },
    {
      "epoch": 0.34,
      "eval_loss": 0.3829396665096283,
      "eval_runtime": 7.3321,
      "eval_samples_per_second": 136.523,
      "eval_steps_per_second": 34.233,
      "step": 775
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.422222222222222e-05,
      "loss": 0.4342,
      "step": 780
    },
    {
      "epoch": 0.35,
      "learning_rate": 4.414814814814815e-05,
      "loss": 0.4007,
      "step": 790
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.4074074074074076e-05,
      "loss": 0.3931,
      "step": 800
    },
    {
      "epoch": 0.36,
      "eval_loss": 0.37909042835235596,
      "eval_runtime": 7.3121,
      "eval_samples_per_second": 136.896,
      "eval_steps_per_second": 34.327,
      "step": 800
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3937,
      "step": 810
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.392592592592593e-05,
      "loss": 0.3655,
      "step": 820
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.37680765986442566,
      "eval_runtime": 7.2101,
      "eval_samples_per_second": 138.832,
      "eval_steps_per_second": 34.812,
      "step": 825
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.385185185185185e-05,
      "loss": 0.3454,
      "step": 830
    },
    {
      "epoch": 0.37,
      "learning_rate": 4.377777777777778e-05,
      "loss": 0.3279,
      "step": 840
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.3703703703703705e-05,
      "loss": 0.4125,
      "step": 850
    },
    {
      "epoch": 0.38,
      "eval_loss": 0.3775666058063507,
      "eval_runtime": 7.2081,
      "eval_samples_per_second": 138.872,
      "eval_steps_per_second": 34.822,
      "step": 850
    },
    {
      "epoch": 0.38,
      "learning_rate": 4.3629629629629635e-05,
      "loss": 0.3324,
      "step": 860
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.355555555555556e-05,
      "loss": 0.3848,
      "step": 870
    },
    {
      "epoch": 0.39,
      "eval_loss": 0.373475044965744,
      "eval_runtime": 7.1966,
      "eval_samples_per_second": 139.093,
      "eval_steps_per_second": 34.877,
      "step": 875
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.348148148148148e-05,
      "loss": 0.5328,
      "step": 880
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.340740740740741e-05,
      "loss": 0.3791,
      "step": 890
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.3907,
      "step": 900
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.3697744607925415,
      "eval_runtime": 7.2557,
      "eval_samples_per_second": 137.961,
      "eval_steps_per_second": 34.594,
      "step": 900
    },
    {
      "epoch": 0.4,
      "learning_rate": 4.325925925925926e-05,
      "loss": 0.3595,
      "step": 910
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.318518518518519e-05,
      "loss": 0.4298,
      "step": 920
    },
    {
      "epoch": 0.41,
      "eval_loss": 0.3690047562122345,
      "eval_runtime": 7.1998,
      "eval_samples_per_second": 139.032,
      "eval_steps_per_second": 34.862,
      "step": 925
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.311111111111111e-05,
      "loss": 0.368,
      "step": 930
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.303703703703704e-05,
      "loss": 0.384,
      "step": 940
    },
    {
      "epoch": 0.42,
      "learning_rate": 4.296296296296296e-05,
      "loss": 0.4369,
      "step": 950
    },
    {
      "epoch": 0.42,
      "eval_loss": 0.3619600534439087,
      "eval_runtime": 7.2818,
      "eval_samples_per_second": 137.467,
      "eval_steps_per_second": 34.47,
      "step": 950
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.2888888888888886e-05,
      "loss": 0.3569,
      "step": 960
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.2814814814814816e-05,
      "loss": 0.3332,
      "step": 970
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.3631249666213989,
      "eval_runtime": 7.2547,
      "eval_samples_per_second": 137.98,
      "eval_steps_per_second": 34.598,
      "step": 975
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.274074074074074e-05,
      "loss": 0.3758,
      "step": 980
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.266666666666667e-05,
      "loss": 0.385,
      "step": 990
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.259259259259259e-05,
      "loss": 0.4328,
      "step": 1000
    },
    {
      "epoch": 0.44,
      "eval_loss": 0.35931944847106934,
      "eval_runtime": 7.3342,
      "eval_samples_per_second": 136.484,
      "eval_steps_per_second": 34.223,
      "step": 1000
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.2518518518518515e-05,
      "loss": 0.4093,
      "step": 1010
    },
    {
      "epoch": 0.45,
      "learning_rate": 4.2444444444444445e-05,
      "loss": 0.3752,
      "step": 1020
    },
    {
      "epoch": 0.46,
      "eval_loss": 0.3549325466156006,
      "eval_runtime": 7.1796,
      "eval_samples_per_second": 139.424,
      "eval_steps_per_second": 34.96,
      "step": 1025
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.237037037037037e-05,
      "loss": 0.3473,
      "step": 1030
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.22962962962963e-05,
      "loss": 0.3823,
      "step": 1040
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.222222222222222e-05,
      "loss": 0.4004,
      "step": 1050
    },
    {
      "epoch": 0.47,
      "eval_loss": 0.3550175428390503,
      "eval_runtime": 7.3,
      "eval_samples_per_second": 137.122,
      "eval_steps_per_second": 34.383,
      "step": 1050
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.2148148148148145e-05,
      "loss": 0.3484,
      "step": 1060
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.2074074074074075e-05,
      "loss": 0.3552,
      "step": 1070
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.3549079895019531,
      "eval_runtime": 7.6065,
      "eval_samples_per_second": 131.599,
      "eval_steps_per_second": 32.998,
      "step": 1075
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.2e-05,
      "loss": 0.3847,
      "step": 1080
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.192592592592593e-05,
      "loss": 0.3861,
      "step": 1090
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.185185185185185e-05,
      "loss": 0.3719,
      "step": 1100
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.3503241539001465,
      "eval_runtime": 7.7017,
      "eval_samples_per_second": 129.971,
      "eval_steps_per_second": 32.59,
      "step": 1100
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.177777777777778e-05,
      "loss": 0.3661,
      "step": 1110
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.1703703703703704e-05,
      "loss": 0.3762,
      "step": 1120
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.3482719659805298,
      "eval_runtime": 7.5987,
      "eval_samples_per_second": 131.732,
      "eval_steps_per_second": 33.032,
      "step": 1125
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.162962962962963e-05,
      "loss": 0.3803,
      "step": 1130
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.155555555555556e-05,
      "loss": 0.2975,
      "step": 1140
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.148148148148148e-05,
      "loss": 0.3397,
      "step": 1150
    },
    {
      "epoch": 0.51,
      "eval_loss": 0.34838753938674927,
      "eval_runtime": 7.5913,
      "eval_samples_per_second": 131.861,
      "eval_steps_per_second": 33.064,
      "step": 1150
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.140740740740741e-05,
      "loss": 0.3218,
      "step": 1160
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.133333333333333e-05,
      "loss": 0.345,
      "step": 1170
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.3447699248790741,
      "eval_runtime": 7.6124,
      "eval_samples_per_second": 131.496,
      "eval_steps_per_second": 32.973,
      "step": 1175
    },
    {
      "epoch": 0.52,
      "learning_rate": 4.1259259259259256e-05,
      "loss": 0.3635,
      "step": 1180
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.1185185185185186e-05,
      "loss": 0.3337,
      "step": 1190
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.111111111111111e-05,
      "loss": 0.3892,
      "step": 1200
    },
    {
      "epoch": 0.53,
      "eval_loss": 0.3436849117279053,
      "eval_runtime": 7.6409,
      "eval_samples_per_second": 131.006,
      "eval_steps_per_second": 32.85,
      "step": 1200
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.103703703703704e-05,
      "loss": 0.3792,
      "step": 1210
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.096296296296296e-05,
      "loss": 0.3062,
      "step": 1220
    },
    {
      "epoch": 0.54,
      "eval_loss": 0.34092003107070923,
      "eval_runtime": 7.6331,
      "eval_samples_per_second": 131.14,
      "eval_steps_per_second": 32.883,
      "step": 1225
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.088888888888889e-05,
      "loss": 0.3764,
      "step": 1230
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.0814814814814815e-05,
      "loss": 0.3541,
      "step": 1240
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.074074074074074e-05,
      "loss": 0.3728,
      "step": 1250
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.3405691683292389,
      "eval_runtime": 7.5692,
      "eval_samples_per_second": 132.247,
      "eval_steps_per_second": 33.161,
      "step": 1250
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.066666666666667e-05,
      "loss": 0.3013,
      "step": 1260
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.059259259259259e-05,
      "loss": 0.321,
      "step": 1270
    },
    {
      "epoch": 0.57,
      "eval_loss": 0.3399699926376343,
      "eval_runtime": 7.588,
      "eval_samples_per_second": 131.919,
      "eval_steps_per_second": 33.079,
      "step": 1275
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.051851851851852e-05,
      "loss": 0.3404,
      "step": 1280
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.0444444444444444e-05,
      "loss": 0.381,
      "step": 1290
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.0370370370370374e-05,
      "loss": 0.314,
      "step": 1300
    },
    {
      "epoch": 0.58,
      "eval_loss": 0.3377434313297272,
      "eval_runtime": 7.5595,
      "eval_samples_per_second": 132.417,
      "eval_steps_per_second": 33.203,
      "step": 1300
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.02962962962963e-05,
      "loss": 0.3016,
      "step": 1310
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.022222222222222e-05,
      "loss": 0.4416,
      "step": 1320
    },
    {
      "epoch": 0.59,
      "eval_loss": 0.33633822202682495,
      "eval_runtime": 7.5093,
      "eval_samples_per_second": 133.302,
      "eval_steps_per_second": 33.425,
      "step": 1325
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.014814814814815e-05,
      "loss": 0.348,
      "step": 1330
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.007407407407407e-05,
      "loss": 0.3685,
      "step": 1340
    },
    {
      "epoch": 0.6,
      "learning_rate": 4e-05,
      "loss": 0.362,
      "step": 1350
    },
    {
      "epoch": 0.6,
      "eval_loss": 0.33278197050094604,
      "eval_runtime": 7.5483,
      "eval_samples_per_second": 132.612,
      "eval_steps_per_second": 33.252,
      "step": 1350
    },
    {
      "epoch": 0.6,
      "learning_rate": 3.9925925925925926e-05,
      "loss": 0.3961,
      "step": 1360
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.985185185185185e-05,
      "loss": 0.2903,
      "step": 1370
    },
    {
      "epoch": 0.61,
      "eval_loss": 0.33177217841148376,
      "eval_runtime": 7.574,
      "eval_samples_per_second": 132.162,
      "eval_steps_per_second": 33.14,
      "step": 1375
    },
    {
      "epoch": 0.61,
      "learning_rate": 3.977777777777778e-05,
      "loss": 0.3527,
      "step": 1380
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.97037037037037e-05,
      "loss": 0.3297,
      "step": 1390
    },
    {
      "epoch": 0.62,
      "learning_rate": 3.962962962962963e-05,
      "loss": 0.3362,
      "step": 1400
    },
    {
      "epoch": 0.62,
      "eval_loss": 0.3336848020553589,
      "eval_runtime": 7.5409,
      "eval_samples_per_second": 132.743,
      "eval_steps_per_second": 33.285,
      "step": 1400
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.9555555555555556e-05,
      "loss": 0.3714,
      "step": 1410
    },
    {
      "epoch": 0.63,
      "learning_rate": 3.9481481481481485e-05,
      "loss": 0.3278,
      "step": 1420
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.32720068097114563,
      "eval_runtime": 7.5879,
      "eval_samples_per_second": 131.921,
      "eval_steps_per_second": 33.079,
      "step": 1425
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.940740740740741e-05,
      "loss": 0.3328,
      "step": 1430
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.933333333333333e-05,
      "loss": 0.2908,
      "step": 1440
    },
    {
      "epoch": 0.64,
      "learning_rate": 3.925925925925926e-05,
      "loss": 0.358,
      "step": 1450
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.3258882761001587,
      "eval_runtime": 7.6673,
      "eval_samples_per_second": 130.554,
      "eval_steps_per_second": 32.736,
      "step": 1450
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9185185185185185e-05,
      "loss": 0.4164,
      "step": 1460
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.9111111111111115e-05,
      "loss": 0.3103,
      "step": 1470
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.3261792063713074,
      "eval_runtime": 7.6042,
      "eval_samples_per_second": 131.637,
      "eval_steps_per_second": 33.008,
      "step": 1475
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.903703703703704e-05,
      "loss": 0.4174,
      "step": 1480
    },
    {
      "epoch": 0.66,
      "learning_rate": 3.896296296296296e-05,
      "loss": 0.3545,
      "step": 1490
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.3297,
      "step": 1500
    },
    {
      "epoch": 0.67,
      "eval_loss": 0.32423877716064453,
      "eval_runtime": 7.6588,
      "eval_samples_per_second": 130.7,
      "eval_steps_per_second": 32.773,
      "step": 1500
    },
    {
      "epoch": 0.67,
      "learning_rate": 3.8814814814814814e-05,
      "loss": 0.3856,
      "step": 1510
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.8740740740740744e-05,
      "loss": 0.3598,
      "step": 1520
    },
    {
      "epoch": 0.68,
      "eval_loss": 0.3228550851345062,
      "eval_runtime": 7.6093,
      "eval_samples_per_second": 131.55,
      "eval_steps_per_second": 32.986,
      "step": 1525
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.866666666666667e-05,
      "loss": 0.3218,
      "step": 1530
    },
    {
      "epoch": 0.68,
      "learning_rate": 3.85925925925926e-05,
      "loss": 0.3577,
      "step": 1540
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.851851851851852e-05,
      "loss": 0.3343,
      "step": 1550
    },
    {
      "epoch": 0.69,
      "eval_loss": 0.32196611166000366,
      "eval_runtime": 7.6896,
      "eval_samples_per_second": 130.175,
      "eval_steps_per_second": 32.641,
      "step": 1550
    },
    {
      "epoch": 0.69,
      "learning_rate": 3.844444444444444e-05,
      "loss": 0.2857,
      "step": 1560
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.837037037037037e-05,
      "loss": 0.3324,
      "step": 1570
    },
    {
      "epoch": 0.7,
      "eval_loss": 0.32020455598831177,
      "eval_runtime": 7.512,
      "eval_samples_per_second": 133.254,
      "eval_steps_per_second": 33.413,
      "step": 1575
    },
    {
      "epoch": 0.7,
      "learning_rate": 3.8296296296296296e-05,
      "loss": 0.3791,
      "step": 1580
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.8222222222222226e-05,
      "loss": 0.3563,
      "step": 1590
    },
    {
      "epoch": 0.71,
      "learning_rate": 3.814814814814815e-05,
      "loss": 0.3744,
      "step": 1600
    },
    {
      "epoch": 0.71,
      "eval_loss": 0.31763964891433716,
      "eval_runtime": 7.5475,
      "eval_samples_per_second": 132.626,
      "eval_steps_per_second": 33.256,
      "step": 1600
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.807407407407408e-05,
      "loss": 0.3689,
      "step": 1610
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.8e-05,
      "loss": 0.3385,
      "step": 1620
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.3172718584537506,
      "eval_runtime": 7.4165,
      "eval_samples_per_second": 134.969,
      "eval_steps_per_second": 33.843,
      "step": 1625
    },
    {
      "epoch": 0.72,
      "learning_rate": 3.7925925925925925e-05,
      "loss": 0.3142,
      "step": 1630
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.7851851851851855e-05,
      "loss": 0.369,
      "step": 1640
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.777777777777778e-05,
      "loss": 0.3264,
      "step": 1650
    },
    {
      "epoch": 0.73,
      "eval_loss": 0.31627902388572693,
      "eval_runtime": 7.4047,
      "eval_samples_per_second": 135.184,
      "eval_steps_per_second": 33.897,
      "step": 1650
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.770370370370371e-05,
      "loss": 0.3481,
      "step": 1660
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.762962962962963e-05,
      "loss": 0.3162,
      "step": 1670
    },
    {
      "epoch": 0.74,
      "eval_loss": 0.31443050503730774,
      "eval_runtime": 7.2703,
      "eval_samples_per_second": 137.683,
      "eval_steps_per_second": 34.524,
      "step": 1675
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7555555555555554e-05,
      "loss": 0.3458,
      "step": 1680
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.7481481481481484e-05,
      "loss": 0.3641,
      "step": 1690
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.740740740740741e-05,
      "loss": 0.3399,
      "step": 1700
    },
    {
      "epoch": 0.76,
      "eval_loss": 0.3139602839946747,
      "eval_runtime": 7.5549,
      "eval_samples_per_second": 132.496,
      "eval_steps_per_second": 33.223,
      "step": 1700
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.733333333333334e-05,
      "loss": 0.3082,
      "step": 1710
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.725925925925926e-05,
      "loss": 0.3544,
      "step": 1720
    },
    {
      "epoch": 0.77,
      "eval_loss": 0.3116389811038971,
      "eval_runtime": 7.3628,
      "eval_samples_per_second": 135.955,
      "eval_steps_per_second": 34.091,
      "step": 1725
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.718518518518519e-05,
      "loss": 0.3356,
      "step": 1730
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.7111111111111113e-05,
      "loss": 0.383,
      "step": 1740
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.7037037037037037e-05,
      "loss": 0.3839,
      "step": 1750
    },
    {
      "epoch": 0.78,
      "eval_loss": 0.3124999403953552,
      "eval_runtime": 7.3086,
      "eval_samples_per_second": 136.962,
      "eval_steps_per_second": 34.343,
      "step": 1750
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.6962962962962966e-05,
      "loss": 0.3378,
      "step": 1760
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.688888888888889e-05,
      "loss": 0.3034,
      "step": 1770
    },
    {
      "epoch": 0.79,
      "eval_loss": 0.31177276372909546,
      "eval_runtime": 7.2443,
      "eval_samples_per_second": 138.177,
      "eval_steps_per_second": 34.648,
      "step": 1775
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.681481481481482e-05,
      "loss": 0.3384,
      "step": 1780
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.674074074074074e-05,
      "loss": 0.3401,
      "step": 1790
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.2989,
      "step": 1800
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.3089355528354645,
      "eval_runtime": 7.1728,
      "eval_samples_per_second": 139.556,
      "eval_steps_per_second": 34.993,
      "step": 1800
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.6592592592592596e-05,
      "loss": 0.3398,
      "step": 1810
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.651851851851852e-05,
      "loss": 0.3055,
      "step": 1820
    },
    {
      "epoch": 0.81,
      "eval_loss": 0.30923065543174744,
      "eval_runtime": 7.208,
      "eval_samples_per_second": 138.874,
      "eval_steps_per_second": 34.822,
      "step": 1825
    },
    {
      "epoch": 0.81,
      "learning_rate": 3.644444444444445e-05,
      "loss": 0.3088,
      "step": 1830
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.637037037037037e-05,
      "loss": 0.3016,
      "step": 1840
    },
    {
      "epoch": 0.82,
      "learning_rate": 3.62962962962963e-05,
      "loss": 0.3143,
      "step": 1850
    },
    {
      "epoch": 0.82,
      "eval_loss": 0.3072822690010071,
      "eval_runtime": 7.2053,
      "eval_samples_per_second": 138.926,
      "eval_steps_per_second": 34.836,
      "step": 1850
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.6222222222222225e-05,
      "loss": 0.2985,
      "step": 1860
    },
    {
      "epoch": 0.83,
      "learning_rate": 3.614814814814815e-05,
      "loss": 0.3805,
      "step": 1870
    },
    {
      "epoch": 0.83,
      "eval_loss": 0.3062894940376282,
      "eval_runtime": 7.2009,
      "eval_samples_per_second": 139.01,
      "eval_steps_per_second": 34.857,
      "step": 1875
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.607407407407408e-05,
      "loss": 0.2831,
      "step": 1880
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.6e-05,
      "loss": 0.3258,
      "step": 1890
    },
    {
      "epoch": 0.84,
      "learning_rate": 3.592592592592593e-05,
      "loss": 0.2625,
      "step": 1900
    },
    {
      "epoch": 0.84,
      "eval_loss": 0.30678972601890564,
      "eval_runtime": 7.1985,
      "eval_samples_per_second": 139.057,
      "eval_steps_per_second": 34.869,
      "step": 1900
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.5851851851851854e-05,
      "loss": 0.2923,
      "step": 1910
    },
    {
      "epoch": 0.85,
      "learning_rate": 3.577777777777778e-05,
      "loss": 0.3276,
      "step": 1920
    },
    {
      "epoch": 0.86,
      "eval_loss": 0.30508390069007874,
      "eval_runtime": 7.2926,
      "eval_samples_per_second": 137.262,
      "eval_steps_per_second": 34.418,
      "step": 1925
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.570370370370371e-05,
      "loss": 0.3521,
      "step": 1930
    },
    {
      "epoch": 0.86,
      "learning_rate": 3.562962962962963e-05,
      "loss": 0.283,
      "step": 1940
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.555555555555556e-05,
      "loss": 0.3364,
      "step": 1950
    },
    {
      "epoch": 0.87,
      "eval_loss": 0.3016437888145447,
      "eval_runtime": 7.3704,
      "eval_samples_per_second": 135.814,
      "eval_steps_per_second": 34.055,
      "step": 1950
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.548148148148148e-05,
      "loss": 0.2882,
      "step": 1960
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.540740740740741e-05,
      "loss": 0.3353,
      "step": 1970
    },
    {
      "epoch": 0.88,
      "eval_loss": 0.30082637071609497,
      "eval_runtime": 7.6161,
      "eval_samples_per_second": 131.432,
      "eval_steps_per_second": 32.957,
      "step": 1975
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.5333333333333336e-05,
      "loss": 0.2675,
      "step": 1980
    },
    {
      "epoch": 0.88,
      "learning_rate": 3.525925925925926e-05,
      "loss": 0.3182,
      "step": 1990
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.518518518518519e-05,
      "loss": 0.2932,
      "step": 2000
    },
    {
      "epoch": 0.89,
      "eval_loss": 0.2999679744243622,
      "eval_runtime": 7.5872,
      "eval_samples_per_second": 131.933,
      "eval_steps_per_second": 33.082,
      "step": 2000
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.511111111111111e-05,
      "loss": 0.3029,
      "step": 2010
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.503703703703704e-05,
      "loss": 0.2434,
      "step": 2020
    },
    {
      "epoch": 0.9,
      "eval_loss": 0.2994518280029297,
      "eval_runtime": 7.5416,
      "eval_samples_per_second": 132.731,
      "eval_steps_per_second": 33.282,
      "step": 2025
    },
    {
      "epoch": 0.9,
      "learning_rate": 3.4962962962962965e-05,
      "loss": 0.3236,
      "step": 2030
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.4888888888888895e-05,
      "loss": 0.2967,
      "step": 2040
    },
    {
      "epoch": 0.91,
      "learning_rate": 3.481481481481482e-05,
      "loss": 0.3572,
      "step": 2050
    },
    {
      "epoch": 0.91,
      "eval_loss": 0.29705023765563965,
      "eval_runtime": 7.5563,
      "eval_samples_per_second": 132.473,
      "eval_steps_per_second": 33.218,
      "step": 2050
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.474074074074074e-05,
      "loss": 0.3384,
      "step": 2060
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.466666666666667e-05,
      "loss": 0.2816,
      "step": 2070
    },
    {
      "epoch": 0.92,
      "eval_loss": 0.29650408029556274,
      "eval_runtime": 7.3924,
      "eval_samples_per_second": 135.409,
      "eval_steps_per_second": 33.954,
      "step": 2075
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.4592592592592594e-05,
      "loss": 0.3361,
      "step": 2080
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.4518518518518524e-05,
      "loss": 0.3125,
      "step": 2090
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.444444444444445e-05,
      "loss": 0.3801,
      "step": 2100
    },
    {
      "epoch": 0.93,
      "eval_loss": 0.2950206696987152,
      "eval_runtime": 7.3773,
      "eval_samples_per_second": 135.687,
      "eval_steps_per_second": 34.023,
      "step": 2100
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.437037037037037e-05,
      "loss": 0.2684,
      "step": 2110
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.42962962962963e-05,
      "loss": 0.3387,
      "step": 2120
    },
    {
      "epoch": 0.94,
      "eval_loss": 0.29483914375305176,
      "eval_runtime": 7.5911,
      "eval_samples_per_second": 131.865,
      "eval_steps_per_second": 33.065,
      "step": 2125
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.4222222222222224e-05,
      "loss": 0.3394,
      "step": 2130
    },
    {
      "epoch": 0.95,
      "learning_rate": 3.4148148148148153e-05,
      "loss": 0.2852,
      "step": 2140
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.4074074074074077e-05,
      "loss": 0.2903,
      "step": 2150
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.2979108989238739,
      "eval_runtime": 7.666,
      "eval_samples_per_second": 130.576,
      "eval_steps_per_second": 32.742,
      "step": 2150
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 0.3144,
      "step": 2160
    },
    {
      "epoch": 0.96,
      "learning_rate": 3.392592592592593e-05,
      "loss": 0.3238,
      "step": 2170
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.2956538200378418,
      "eval_runtime": 7.6479,
      "eval_samples_per_second": 130.886,
      "eval_steps_per_second": 32.82,
      "step": 2175
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.385185185185185e-05,
      "loss": 0.2848,
      "step": 2180
    },
    {
      "epoch": 0.97,
      "learning_rate": 3.377777777777778e-05,
      "loss": 0.2742,
      "step": 2190
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3703703703703706e-05,
      "loss": 0.3392,
      "step": 2200
    },
    {
      "epoch": 0.98,
      "eval_loss": 0.29324308037757874,
      "eval_runtime": 7.6322,
      "eval_samples_per_second": 131.155,
      "eval_steps_per_second": 32.887,
      "step": 2200
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.3629629629629636e-05,
      "loss": 0.2332,
      "step": 2210
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.355555555555556e-05,
      "loss": 0.2754,
      "step": 2220
    },
    {
      "epoch": 0.99,
      "eval_loss": 0.29224133491516113,
      "eval_runtime": 7.5764,
      "eval_samples_per_second": 132.12,
      "eval_steps_per_second": 33.129,
      "step": 2225
    },
    {
      "epoch": 0.99,
      "learning_rate": 3.348148148148148e-05,
      "loss": 0.2778,
      "step": 2230
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.340740740740741e-05,
      "loss": 0.2646,
      "step": 2240
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.3542,
      "step": 2250
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.2917466461658478,
      "eval_runtime": 7.5812,
      "eval_samples_per_second": 132.038,
      "eval_steps_per_second": 33.108,
      "step": 2250
    },
    {
      "epoch": 1.0,
      "learning_rate": 3.3259259259259265e-05,
      "loss": 0.3091,
      "step": 2260
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.318518518518519e-05,
      "loss": 0.314,
      "step": 2270
    },
    {
      "epoch": 1.01,
      "eval_loss": 0.29272007942199707,
      "eval_runtime": 7.5219,
      "eval_samples_per_second": 133.079,
      "eval_steps_per_second": 33.369,
      "step": 2275
    },
    {
      "epoch": 1.01,
      "learning_rate": 3.311111111111112e-05,
      "loss": 0.269,
      "step": 2280
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.303703703703704e-05,
      "loss": 0.2504,
      "step": 2290
    },
    {
      "epoch": 1.02,
      "learning_rate": 3.2962962962962964e-05,
      "loss": 0.3201,
      "step": 2300
    },
    {
      "epoch": 1.02,
      "eval_loss": 0.29086679220199585,
      "eval_runtime": 7.5125,
      "eval_samples_per_second": 133.245,
      "eval_steps_per_second": 33.411,
      "step": 2300
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.2888888888888894e-05,
      "loss": 0.2967,
      "step": 2310
    },
    {
      "epoch": 1.03,
      "learning_rate": 3.281481481481482e-05,
      "loss": 0.2799,
      "step": 2320
    },
    {
      "epoch": 1.03,
      "eval_loss": 0.28955137729644775,
      "eval_runtime": 7.3165,
      "eval_samples_per_second": 136.814,
      "eval_steps_per_second": 34.306,
      "step": 2325
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.274074074074075e-05,
      "loss": 0.2535,
      "step": 2330
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.266666666666667e-05,
      "loss": 0.2763,
      "step": 2340
    },
    {
      "epoch": 1.04,
      "learning_rate": 3.25925925925926e-05,
      "loss": 0.2356,
      "step": 2350
    },
    {
      "epoch": 1.04,
      "eval_loss": 0.2900914251804352,
      "eval_runtime": 7.3036,
      "eval_samples_per_second": 137.055,
      "eval_steps_per_second": 34.366,
      "step": 2350
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.251851851851852e-05,
      "loss": 0.3099,
      "step": 2360
    },
    {
      "epoch": 1.05,
      "learning_rate": 3.2444444444444446e-05,
      "loss": 0.271,
      "step": 2370
    },
    {
      "epoch": 1.06,
      "eval_loss": 0.28874123096466064,
      "eval_runtime": 7.2862,
      "eval_samples_per_second": 137.382,
      "eval_steps_per_second": 34.448,
      "step": 2375
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.2370370370370376e-05,
      "loss": 0.32,
      "step": 2380
    },
    {
      "epoch": 1.06,
      "learning_rate": 3.22962962962963e-05,
      "loss": 0.28,
      "step": 2390
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.222222222222223e-05,
      "loss": 0.2246,
      "step": 2400
    },
    {
      "epoch": 1.07,
      "eval_loss": 0.2897484004497528,
      "eval_runtime": 7.2105,
      "eval_samples_per_second": 138.825,
      "eval_steps_per_second": 34.81,
      "step": 2400
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.214814814814815e-05,
      "loss": 0.3076,
      "step": 2410
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.2074074074074075e-05,
      "loss": 0.2824,
      "step": 2420
    },
    {
      "epoch": 1.08,
      "eval_loss": 0.2881883680820465,
      "eval_runtime": 7.5403,
      "eval_samples_per_second": 132.752,
      "eval_steps_per_second": 33.288,
      "step": 2425
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 0.3322,
      "step": 2430
    },
    {
      "epoch": 1.08,
      "learning_rate": 3.192592592592593e-05,
      "loss": 0.2299,
      "step": 2440
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.185185185185185e-05,
      "loss": 0.2516,
      "step": 2450
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.287548303604126,
      "eval_runtime": 7.3579,
      "eval_samples_per_second": 136.045,
      "eval_steps_per_second": 34.113,
      "step": 2450
    },
    {
      "epoch": 1.09,
      "learning_rate": 3.177777777777778e-05,
      "loss": 0.2652,
      "step": 2460
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.1703703703703705e-05,
      "loss": 0.2465,
      "step": 2470
    },
    {
      "epoch": 1.1,
      "eval_loss": 0.2856321632862091,
      "eval_runtime": 7.7527,
      "eval_samples_per_second": 129.116,
      "eval_steps_per_second": 32.376,
      "step": 2475
    },
    {
      "epoch": 1.1,
      "learning_rate": 3.1629629629629634e-05,
      "loss": 0.301,
      "step": 2480
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.155555555555556e-05,
      "loss": 0.1987,
      "step": 2490
    },
    {
      "epoch": 1.11,
      "learning_rate": 3.148148148148148e-05,
      "loss": 0.3417,
      "step": 2500
    },
    {
      "epoch": 1.11,
      "eval_loss": 0.2860187292098999,
      "eval_runtime": 7.5819,
      "eval_samples_per_second": 132.024,
      "eval_steps_per_second": 33.105,
      "step": 2500
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.140740740740741e-05,
      "loss": 0.2464,
      "step": 2510
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.1333333333333334e-05,
      "loss": 0.2418,
      "step": 2520
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.28458064794540405,
      "eval_runtime": 7.684,
      "eval_samples_per_second": 130.271,
      "eval_steps_per_second": 32.665,
      "step": 2525
    },
    {
      "epoch": 1.12,
      "learning_rate": 3.1259259259259264e-05,
      "loss": 0.2884,
      "step": 2530
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.118518518518519e-05,
      "loss": 0.2396,
      "step": 2540
    },
    {
      "epoch": 1.13,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.2625,
      "step": 2550
    },
    {
      "epoch": 1.13,
      "eval_loss": 0.28442564606666565,
      "eval_runtime": 7.6251,
      "eval_samples_per_second": 131.277,
      "eval_steps_per_second": 32.918,
      "step": 2550
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.103703703703704e-05,
      "loss": 0.2286,
      "step": 2560
    },
    {
      "epoch": 1.14,
      "learning_rate": 3.096296296296296e-05,
      "loss": 0.3023,
      "step": 2570
    },
    {
      "epoch": 1.14,
      "eval_loss": 0.28357136249542236,
      "eval_runtime": 7.6709,
      "eval_samples_per_second": 130.493,
      "eval_steps_per_second": 32.721,
      "step": 2575
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.088888888888889e-05,
      "loss": 0.2631,
      "step": 2580
    },
    {
      "epoch": 1.15,
      "learning_rate": 3.0814814814814816e-05,
      "loss": 0.2656,
      "step": 2590
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.074074074074074e-05,
      "loss": 0.2301,
      "step": 2600
    },
    {
      "epoch": 1.16,
      "eval_loss": 0.2838016450405121,
      "eval_runtime": 7.4972,
      "eval_samples_per_second": 133.516,
      "eval_steps_per_second": 33.479,
      "step": 2600
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.066666666666667e-05,
      "loss": 0.2708,
      "step": 2610
    },
    {
      "epoch": 1.16,
      "learning_rate": 3.059259259259259e-05,
      "loss": 0.2638,
      "step": 2620
    },
    {
      "epoch": 1.17,
      "eval_loss": 0.2819526791572571,
      "eval_runtime": 7.5414,
      "eval_samples_per_second": 132.734,
      "eval_steps_per_second": 33.283,
      "step": 2625
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.0518518518518515e-05,
      "loss": 0.2786,
      "step": 2630
    },
    {
      "epoch": 1.17,
      "learning_rate": 3.044444444444445e-05,
      "loss": 0.3328,
      "step": 2640
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.037037037037037e-05,
      "loss": 0.2835,
      "step": 2650
    },
    {
      "epoch": 1.18,
      "eval_loss": 0.2829004228115082,
      "eval_runtime": 7.3311,
      "eval_samples_per_second": 136.541,
      "eval_steps_per_second": 34.238,
      "step": 2650
    },
    {
      "epoch": 1.18,
      "learning_rate": 3.02962962962963e-05,
      "loss": 0.3523,
      "step": 2660
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0222222222222225e-05,
      "loss": 0.2512,
      "step": 2670
    },
    {
      "epoch": 1.19,
      "eval_loss": 0.28273966908454895,
      "eval_runtime": 7.3473,
      "eval_samples_per_second": 136.24,
      "eval_steps_per_second": 34.162,
      "step": 2675
    },
    {
      "epoch": 1.19,
      "learning_rate": 3.0148148148148148e-05,
      "loss": 0.2382,
      "step": 2680
    },
    {
      "epoch": 1.2,
      "learning_rate": 3.0074074074074078e-05,
      "loss": 0.2668,
      "step": 2690
    },
    {
      "epoch": 1.2,
      "learning_rate": 3e-05,
      "loss": 0.2472,
      "step": 2700
    },
    {
      "epoch": 1.2,
      "eval_loss": 0.28208473324775696,
      "eval_runtime": 7.2579,
      "eval_samples_per_second": 137.92,
      "eval_steps_per_second": 34.583,
      "step": 2700
    },
    {
      "epoch": 1.2,
      "learning_rate": 2.992592592592593e-05,
      "loss": 0.2332,
      "step": 2710
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9851851851851854e-05,
      "loss": 0.266,
      "step": 2720
    },
    {
      "epoch": 1.21,
      "eval_loss": 0.281426340341568,
      "eval_runtime": 7.2593,
      "eval_samples_per_second": 137.892,
      "eval_steps_per_second": 34.576,
      "step": 2725
    },
    {
      "epoch": 1.21,
      "learning_rate": 2.9777777777777777e-05,
      "loss": 0.2908,
      "step": 2730
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.9703703703703707e-05,
      "loss": 0.2299,
      "step": 2740
    },
    {
      "epoch": 1.22,
      "learning_rate": 2.962962962962963e-05,
      "loss": 0.2284,
      "step": 2750
    },
    {
      "epoch": 1.22,
      "eval_loss": 0.2816203832626343,
      "eval_runtime": 7.2174,
      "eval_samples_per_second": 138.694,
      "eval_steps_per_second": 34.777,
      "step": 2750
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.955555555555556e-05,
      "loss": 0.2821,
      "step": 2760
    },
    {
      "epoch": 1.23,
      "learning_rate": 2.9481481481481483e-05,
      "loss": 0.3079,
      "step": 2770
    },
    {
      "epoch": 1.23,
      "eval_loss": 0.2785053551197052,
      "eval_runtime": 7.3,
      "eval_samples_per_second": 137.124,
      "eval_steps_per_second": 34.384,
      "step": 2775
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9407407407407413e-05,
      "loss": 0.1826,
      "step": 2780
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.9333333333333336e-05,
      "loss": 0.2438,
      "step": 2790
    },
    {
      "epoch": 1.24,
      "learning_rate": 2.925925925925926e-05,
      "loss": 0.2014,
      "step": 2800
    },
    {
      "epoch": 1.24,
      "eval_loss": 0.2803582549095154,
      "eval_runtime": 7.196,
      "eval_samples_per_second": 139.106,
      "eval_steps_per_second": 34.881,
      "step": 2800
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.918518518518519e-05,
      "loss": 0.2979,
      "step": 2810
    },
    {
      "epoch": 1.25,
      "learning_rate": 2.9111111111111112e-05,
      "loss": 0.2223,
      "step": 2820
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.27921053767204285,
      "eval_runtime": 7.2996,
      "eval_samples_per_second": 137.132,
      "eval_steps_per_second": 34.386,
      "step": 2825
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.9037037037037042e-05,
      "loss": 0.2515,
      "step": 2830
    },
    {
      "epoch": 1.26,
      "learning_rate": 2.8962962962962965e-05,
      "loss": 0.2771,
      "step": 2840
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 0.2611,
      "step": 2850
    },
    {
      "epoch": 1.27,
      "eval_loss": 0.2792360186576843,
      "eval_runtime": 7.1956,
      "eval_samples_per_second": 139.112,
      "eval_steps_per_second": 34.882,
      "step": 2850
    },
    {
      "epoch": 1.27,
      "learning_rate": 2.8814814814814818e-05,
      "loss": 0.2867,
      "step": 2860
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.874074074074074e-05,
      "loss": 0.2492,
      "step": 2870
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.2788338363170624,
      "eval_runtime": 7.2868,
      "eval_samples_per_second": 137.371,
      "eval_steps_per_second": 34.446,
      "step": 2875
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.8666666666666668e-05,
      "loss": 0.26,
      "step": 2880
    },
    {
      "epoch": 1.28,
      "learning_rate": 2.8592592592592594e-05,
      "loss": 0.3693,
      "step": 2890
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.851851851851852e-05,
      "loss": 0.2686,
      "step": 2900
    },
    {
      "epoch": 1.29,
      "eval_loss": 0.27813464403152466,
      "eval_runtime": 7.1901,
      "eval_samples_per_second": 139.219,
      "eval_steps_per_second": 34.909,
      "step": 2900
    },
    {
      "epoch": 1.29,
      "learning_rate": 2.8444444444444447e-05,
      "loss": 0.2091,
      "step": 2910
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.837037037037037e-05,
      "loss": 0.2701,
      "step": 2920
    },
    {
      "epoch": 1.3,
      "eval_loss": 0.27804136276245117,
      "eval_runtime": 7.2533,
      "eval_samples_per_second": 138.006,
      "eval_steps_per_second": 34.605,
      "step": 2925
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.8296296296296297e-05,
      "loss": 0.2795,
      "step": 2930
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.8222222222222223e-05,
      "loss": 0.3346,
      "step": 2940
    },
    {
      "epoch": 1.31,
      "learning_rate": 2.814814814814815e-05,
      "loss": 0.2254,
      "step": 2950
    },
    {
      "epoch": 1.31,
      "eval_loss": 0.27507713437080383,
      "eval_runtime": 7.2159,
      "eval_samples_per_second": 138.721,
      "eval_steps_per_second": 34.784,
      "step": 2950
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.8074074074074076e-05,
      "loss": 0.2507,
      "step": 2960
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 0.2942,
      "step": 2970
    },
    {
      "epoch": 1.32,
      "eval_loss": 0.2757761776447296,
      "eval_runtime": 7.2423,
      "eval_samples_per_second": 138.216,
      "eval_steps_per_second": 34.658,
      "step": 2975
    },
    {
      "epoch": 1.32,
      "learning_rate": 2.7925925925925926e-05,
      "loss": 0.2678,
      "step": 2980
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.7851851851851853e-05,
      "loss": 0.2459,
      "step": 2990
    },
    {
      "epoch": 1.33,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.2446,
      "step": 3000
    },
    {
      "epoch": 1.33,
      "eval_loss": 0.2748652994632721,
      "eval_runtime": 7.2844,
      "eval_samples_per_second": 137.416,
      "eval_steps_per_second": 34.457,
      "step": 3000
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7703703703703706e-05,
      "loss": 0.2278,
      "step": 3010
    },
    {
      "epoch": 1.34,
      "learning_rate": 2.7629629629629632e-05,
      "loss": 0.2337,
      "step": 3020
    },
    {
      "epoch": 1.34,
      "eval_loss": 0.2740112841129303,
      "eval_runtime": 7.3832,
      "eval_samples_per_second": 135.578,
      "eval_steps_per_second": 33.996,
      "step": 3025
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.7555555555555555e-05,
      "loss": 0.2335,
      "step": 3030
    },
    {
      "epoch": 1.35,
      "learning_rate": 2.7481481481481482e-05,
      "loss": 0.2023,
      "step": 3040
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.7407407407407408e-05,
      "loss": 0.2166,
      "step": 3050
    },
    {
      "epoch": 1.36,
      "eval_loss": 0.27378755807876587,
      "eval_runtime": 7.4184,
      "eval_samples_per_second": 134.934,
      "eval_steps_per_second": 33.835,
      "step": 3050
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.733333333333333e-05,
      "loss": 0.2161,
      "step": 3060
    },
    {
      "epoch": 1.36,
      "learning_rate": 2.725925925925926e-05,
      "loss": 0.3003,
      "step": 3070
    },
    {
      "epoch": 1.37,
      "eval_loss": 0.2731979191303253,
      "eval_runtime": 13.3786,
      "eval_samples_per_second": 74.821,
      "eval_steps_per_second": 18.761,
      "step": 3075
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.7185185185185184e-05,
      "loss": 0.2498,
      "step": 3080
    },
    {
      "epoch": 1.37,
      "learning_rate": 2.7111111111111114e-05,
      "loss": 0.2749,
      "step": 3090
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.7037037037037037e-05,
      "loss": 0.2422,
      "step": 3100
    },
    {
      "epoch": 1.38,
      "eval_loss": 0.27196845412254333,
      "eval_runtime": 7.5816,
      "eval_samples_per_second": 132.03,
      "eval_steps_per_second": 33.107,
      "step": 3100
    },
    {
      "epoch": 1.38,
      "learning_rate": 2.696296296296296e-05,
      "loss": 0.2468,
      "step": 3110
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.688888888888889e-05,
      "loss": 0.263,
      "step": 3120
    },
    {
      "epoch": 1.39,
      "eval_loss": 0.2727610766887665,
      "eval_runtime": 7.6161,
      "eval_samples_per_second": 131.432,
      "eval_steps_per_second": 32.956,
      "step": 3125
    },
    {
      "epoch": 1.39,
      "learning_rate": 2.6814814814814814e-05,
      "loss": 0.2436,
      "step": 3130
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.6740740740740743e-05,
      "loss": 0.27,
      "step": 3140
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.2462,
      "step": 3150
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.2711770832538605,
      "eval_runtime": 7.4596,
      "eval_samples_per_second": 134.19,
      "eval_steps_per_second": 33.648,
      "step": 3150
    },
    {
      "epoch": 1.4,
      "learning_rate": 2.659259259259259e-05,
      "loss": 0.3066,
      "step": 3160
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.651851851851852e-05,
      "loss": 0.2262,
      "step": 3170
    },
    {
      "epoch": 1.41,
      "eval_loss": 0.27057451009750366,
      "eval_runtime": 7.4675,
      "eval_samples_per_second": 134.047,
      "eval_steps_per_second": 33.612,
      "step": 3175
    },
    {
      "epoch": 1.41,
      "learning_rate": 2.6444444444444443e-05,
      "loss": 0.2187,
      "step": 3180
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.6370370370370373e-05,
      "loss": 0.2992,
      "step": 3190
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.6296296296296296e-05,
      "loss": 0.3181,
      "step": 3200
    },
    {
      "epoch": 1.42,
      "eval_loss": 0.27069294452667236,
      "eval_runtime": 7.2871,
      "eval_samples_per_second": 137.367,
      "eval_steps_per_second": 34.445,
      "step": 3200
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.6222222222222226e-05,
      "loss": 0.275,
      "step": 3210
    },
    {
      "epoch": 1.43,
      "learning_rate": 2.614814814814815e-05,
      "loss": 0.2337,
      "step": 3220
    },
    {
      "epoch": 1.43,
      "eval_loss": 0.2696068286895752,
      "eval_runtime": 7.3083,
      "eval_samples_per_second": 136.967,
      "eval_steps_per_second": 34.344,
      "step": 3225
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.6074074074074072e-05,
      "loss": 0.2596,
      "step": 3230
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 0.2511,
      "step": 3240
    },
    {
      "epoch": 1.44,
      "learning_rate": 2.5925925925925925e-05,
      "loss": 0.2223,
      "step": 3250
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.26948779821395874,
      "eval_runtime": 7.2594,
      "eval_samples_per_second": 137.89,
      "eval_steps_per_second": 34.576,
      "step": 3250
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.5851851851851855e-05,
      "loss": 0.266,
      "step": 3260
    },
    {
      "epoch": 1.45,
      "learning_rate": 2.5777777777777778e-05,
      "loss": 0.2594,
      "step": 3270
    },
    {
      "epoch": 1.46,
      "eval_loss": 0.2680164873600006,
      "eval_runtime": 7.3879,
      "eval_samples_per_second": 135.492,
      "eval_steps_per_second": 33.975,
      "step": 3275
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.5703703703703708e-05,
      "loss": 0.2623,
      "step": 3280
    },
    {
      "epoch": 1.46,
      "learning_rate": 2.562962962962963e-05,
      "loss": 0.2028,
      "step": 3290
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 0.2226,
      "step": 3300
    },
    {
      "epoch": 1.47,
      "eval_loss": 0.27027028799057007,
      "eval_runtime": 7.2794,
      "eval_samples_per_second": 137.512,
      "eval_steps_per_second": 34.481,
      "step": 3300
    },
    {
      "epoch": 1.47,
      "learning_rate": 2.5481481481481484e-05,
      "loss": 0.211,
      "step": 3310
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5407407407407407e-05,
      "loss": 0.2037,
      "step": 3320
    },
    {
      "epoch": 1.48,
      "eval_loss": 0.2683195471763611,
      "eval_runtime": 7.3164,
      "eval_samples_per_second": 136.815,
      "eval_steps_per_second": 34.306,
      "step": 3325
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.5333333333333337e-05,
      "loss": 0.1988,
      "step": 3330
    },
    {
      "epoch": 1.48,
      "learning_rate": 2.525925925925926e-05,
      "loss": 0.2262,
      "step": 3340
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5185185185185183e-05,
      "loss": 0.2538,
      "step": 3350
    },
    {
      "epoch": 1.49,
      "eval_loss": 0.26811686158180237,
      "eval_runtime": 7.2097,
      "eval_samples_per_second": 138.841,
      "eval_steps_per_second": 34.814,
      "step": 3350
    },
    {
      "epoch": 1.49,
      "learning_rate": 2.5111111111111113e-05,
      "loss": 0.2029,
      "step": 3360
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.5037037037037036e-05,
      "loss": 0.2094,
      "step": 3370
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.269042432308197,
      "eval_runtime": 7.2591,
      "eval_samples_per_second": 137.896,
      "eval_steps_per_second": 34.577,
      "step": 3375
    },
    {
      "epoch": 1.5,
      "learning_rate": 2.4962962962962963e-05,
      "loss": 0.195,
      "step": 3380
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.488888888888889e-05,
      "loss": 0.2639,
      "step": 3390
    },
    {
      "epoch": 1.51,
      "learning_rate": 2.4814814814814816e-05,
      "loss": 0.2591,
      "step": 3400
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.26725488901138306,
      "eval_runtime": 7.2092,
      "eval_samples_per_second": 138.851,
      "eval_steps_per_second": 34.817,
      "step": 3400
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4740740740740742e-05,
      "loss": 0.2043,
      "step": 3410
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.466666666666667e-05,
      "loss": 0.2274,
      "step": 3420
    },
    {
      "epoch": 1.52,
      "eval_loss": 0.2678578794002533,
      "eval_runtime": 7.2753,
      "eval_samples_per_second": 137.588,
      "eval_steps_per_second": 34.5,
      "step": 3425
    },
    {
      "epoch": 1.52,
      "learning_rate": 2.4592592592592595e-05,
      "loss": 0.2352,
      "step": 3430
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.451851851851852e-05,
      "loss": 0.2365,
      "step": 3440
    },
    {
      "epoch": 1.53,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 0.2526,
      "step": 3450
    },
    {
      "epoch": 1.53,
      "eval_loss": 0.266437292098999,
      "eval_runtime": 7.2076,
      "eval_samples_per_second": 138.88,
      "eval_steps_per_second": 34.824,
      "step": 3450
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.437037037037037e-05,
      "loss": 0.1861,
      "step": 3460
    },
    {
      "epoch": 1.54,
      "learning_rate": 2.4296296296296298e-05,
      "loss": 0.2659,
      "step": 3470
    },
    {
      "epoch": 1.54,
      "eval_loss": 0.265876829624176,
      "eval_runtime": 7.5122,
      "eval_samples_per_second": 133.25,
      "eval_steps_per_second": 33.412,
      "step": 3475
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.4222222222222224e-05,
      "loss": 0.2159,
      "step": 3480
    },
    {
      "epoch": 1.55,
      "learning_rate": 2.414814814814815e-05,
      "loss": 0.2809,
      "step": 3490
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.4074074074074074e-05,
      "loss": 0.2144,
      "step": 3500
    },
    {
      "epoch": 1.56,
      "eval_loss": 0.2642614245414734,
      "eval_runtime": 7.468,
      "eval_samples_per_second": 134.039,
      "eval_steps_per_second": 33.61,
      "step": 3500
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.4e-05,
      "loss": 0.2122,
      "step": 3510
    },
    {
      "epoch": 1.56,
      "learning_rate": 2.3925925925925927e-05,
      "loss": 0.2432,
      "step": 3520
    },
    {
      "epoch": 1.57,
      "eval_loss": 0.2640276551246643,
      "eval_runtime": 7.5832,
      "eval_samples_per_second": 132.003,
      "eval_steps_per_second": 33.1,
      "step": 3525
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.3851851851851854e-05,
      "loss": 0.2441,
      "step": 3530
    },
    {
      "epoch": 1.57,
      "learning_rate": 2.377777777777778e-05,
      "loss": 0.2647,
      "step": 3540
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.3703703703703707e-05,
      "loss": 0.2852,
      "step": 3550
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.26369205117225647,
      "eval_runtime": 7.3907,
      "eval_samples_per_second": 135.44,
      "eval_steps_per_second": 33.962,
      "step": 3550
    },
    {
      "epoch": 1.58,
      "learning_rate": 2.3629629629629633e-05,
      "loss": 0.2321,
      "step": 3560
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3555555555555556e-05,
      "loss": 0.3375,
      "step": 3570
    },
    {
      "epoch": 1.59,
      "eval_loss": 0.2633427381515503,
      "eval_runtime": 7.3881,
      "eval_samples_per_second": 135.489,
      "eval_steps_per_second": 33.974,
      "step": 3575
    },
    {
      "epoch": 1.59,
      "learning_rate": 2.3481481481481483e-05,
      "loss": 0.2219,
      "step": 3580
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.340740740740741e-05,
      "loss": 0.2004,
      "step": 3590
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.272,
      "step": 3600
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.26376578211784363,
      "eval_runtime": 7.2236,
      "eval_samples_per_second": 138.574,
      "eval_steps_per_second": 34.747,
      "step": 3600
    },
    {
      "epoch": 1.6,
      "learning_rate": 2.3259259259259262e-05,
      "loss": 0.2887,
      "step": 3610
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.318518518518519e-05,
      "loss": 0.2502,
      "step": 3620
    },
    {
      "epoch": 1.61,
      "eval_loss": 0.2631487250328064,
      "eval_runtime": 7.2851,
      "eval_samples_per_second": 137.403,
      "eval_steps_per_second": 34.454,
      "step": 3625
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.3111111111111112e-05,
      "loss": 0.206,
      "step": 3630
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.303703703703704e-05,
      "loss": 0.2368,
      "step": 3640
    },
    {
      "epoch": 1.62,
      "learning_rate": 2.2962962962962965e-05,
      "loss": 0.2131,
      "step": 3650
    },
    {
      "epoch": 1.62,
      "eval_loss": 0.2649364173412323,
      "eval_runtime": 7.1872,
      "eval_samples_per_second": 139.275,
      "eval_steps_per_second": 34.923,
      "step": 3650
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.288888888888889e-05,
      "loss": 0.2137,
      "step": 3660
    },
    {
      "epoch": 1.63,
      "learning_rate": 2.2814814814814818e-05,
      "loss": 0.2313,
      "step": 3670
    },
    {
      "epoch": 1.63,
      "eval_loss": 0.2636246979236603,
      "eval_runtime": 7.2744,
      "eval_samples_per_second": 137.606,
      "eval_steps_per_second": 34.505,
      "step": 3675
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.2740740740740744e-05,
      "loss": 0.239,
      "step": 3680
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.2666666666666668e-05,
      "loss": 0.2354,
      "step": 3690
    },
    {
      "epoch": 1.64,
      "learning_rate": 2.2592592592592594e-05,
      "loss": 0.1919,
      "step": 3700
    },
    {
      "epoch": 1.64,
      "eval_loss": 0.262579083442688,
      "eval_runtime": 7.2981,
      "eval_samples_per_second": 137.159,
      "eval_steps_per_second": 34.392,
      "step": 3700
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.251851851851852e-05,
      "loss": 0.2836,
      "step": 3710
    },
    {
      "epoch": 1.65,
      "learning_rate": 2.2444444444444447e-05,
      "loss": 0.2761,
      "step": 3720
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.26134076714515686,
      "eval_runtime": 7.3981,
      "eval_samples_per_second": 135.305,
      "eval_steps_per_second": 33.928,
      "step": 3725
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.2370370370370374e-05,
      "loss": 0.1932,
      "step": 3730
    },
    {
      "epoch": 1.66,
      "learning_rate": 2.2296296296296297e-05,
      "loss": 0.2248,
      "step": 3740
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.2151,
      "step": 3750
    },
    {
      "epoch": 1.67,
      "eval_loss": 0.26156380772590637,
      "eval_runtime": 7.2684,
      "eval_samples_per_second": 137.72,
      "eval_steps_per_second": 34.533,
      "step": 3750
    },
    {
      "epoch": 1.67,
      "learning_rate": 2.214814814814815e-05,
      "loss": 0.2614,
      "step": 3760
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.2074074074074076e-05,
      "loss": 0.2725,
      "step": 3770
    },
    {
      "epoch": 1.68,
      "eval_loss": 0.26058655977249146,
      "eval_runtime": 7.2639,
      "eval_samples_per_second": 137.804,
      "eval_steps_per_second": 34.554,
      "step": 3775
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 0.2402,
      "step": 3780
    },
    {
      "epoch": 1.68,
      "learning_rate": 2.1925925925925926e-05,
      "loss": 0.2283,
      "step": 3790
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.1851851851851852e-05,
      "loss": 0.2552,
      "step": 3800
    },
    {
      "epoch": 1.69,
      "eval_loss": 0.26024872064590454,
      "eval_runtime": 7.191,
      "eval_samples_per_second": 139.201,
      "eval_steps_per_second": 34.905,
      "step": 3800
    },
    {
      "epoch": 1.69,
      "learning_rate": 2.177777777777778e-05,
      "loss": 0.2615,
      "step": 3810
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.1703703703703705e-05,
      "loss": 0.2094,
      "step": 3820
    },
    {
      "epoch": 1.7,
      "eval_loss": 0.2611374258995056,
      "eval_runtime": 7.1864,
      "eval_samples_per_second": 139.291,
      "eval_steps_per_second": 34.927,
      "step": 3825
    },
    {
      "epoch": 1.7,
      "learning_rate": 2.162962962962963e-05,
      "loss": 0.2267,
      "step": 3830
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.1555555555555555e-05,
      "loss": 0.2807,
      "step": 3840
    },
    {
      "epoch": 1.71,
      "learning_rate": 2.148148148148148e-05,
      "loss": 0.2948,
      "step": 3850
    },
    {
      "epoch": 1.71,
      "eval_loss": 0.25876420736312866,
      "eval_runtime": 7.2079,
      "eval_samples_per_second": 138.875,
      "eval_steps_per_second": 34.823,
      "step": 3850
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.1407407407407408e-05,
      "loss": 0.261,
      "step": 3860
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.1333333333333335e-05,
      "loss": 0.3131,
      "step": 3870
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.25907066464424133,
      "eval_runtime": 7.2187,
      "eval_samples_per_second": 138.668,
      "eval_steps_per_second": 34.771,
      "step": 3875
    },
    {
      "epoch": 1.72,
      "learning_rate": 2.1259259259259258e-05,
      "loss": 0.2667,
      "step": 3880
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.1185185185185184e-05,
      "loss": 0.216,
      "step": 3890
    },
    {
      "epoch": 1.73,
      "learning_rate": 2.111111111111111e-05,
      "loss": 0.237,
      "step": 3900
    },
    {
      "epoch": 1.73,
      "eval_loss": 0.2582587003707886,
      "eval_runtime": 8.5503,
      "eval_samples_per_second": 117.072,
      "eval_steps_per_second": 29.356,
      "step": 3900
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.1037037037037037e-05,
      "loss": 0.3102,
      "step": 3910
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.0962962962962964e-05,
      "loss": 0.2385,
      "step": 3920
    },
    {
      "epoch": 1.74,
      "eval_loss": 0.2583344876766205,
      "eval_runtime": 7.2821,
      "eval_samples_per_second": 137.461,
      "eval_steps_per_second": 34.468,
      "step": 3925
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.088888888888889e-05,
      "loss": 0.3235,
      "step": 3930
    },
    {
      "epoch": 1.75,
      "learning_rate": 2.0814814814814813e-05,
      "loss": 0.2154,
      "step": 3940
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.074074074074074e-05,
      "loss": 0.1842,
      "step": 3950
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.2580818831920624,
      "eval_runtime": 7.213,
      "eval_samples_per_second": 138.777,
      "eval_steps_per_second": 34.798,
      "step": 3950
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.0666666666666666e-05,
      "loss": 0.2051,
      "step": 3960
    },
    {
      "epoch": 1.76,
      "learning_rate": 2.0592592592592593e-05,
      "loss": 0.2291,
      "step": 3970
    },
    {
      "epoch": 1.77,
      "eval_loss": 0.25752562284469604,
      "eval_runtime": 7.2125,
      "eval_samples_per_second": 138.787,
      "eval_steps_per_second": 34.801,
      "step": 3975
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.051851851851852e-05,
      "loss": 0.1973,
      "step": 3980
    },
    {
      "epoch": 1.77,
      "learning_rate": 2.0444444444444446e-05,
      "loss": 0.2864,
      "step": 3990
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.037037037037037e-05,
      "loss": 0.2124,
      "step": 4000
    },
    {
      "epoch": 1.78,
      "eval_loss": 0.2576400637626648,
      "eval_runtime": 7.2072,
      "eval_samples_per_second": 138.89,
      "eval_steps_per_second": 34.827,
      "step": 4000
    },
    {
      "epoch": 1.78,
      "learning_rate": 2.0296296296296296e-05,
      "loss": 0.3026,
      "step": 4010
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.0222222222222222e-05,
      "loss": 0.2713,
      "step": 4020
    },
    {
      "epoch": 1.79,
      "eval_loss": 0.256246954202652,
      "eval_runtime": 7.2163,
      "eval_samples_per_second": 138.714,
      "eval_steps_per_second": 34.782,
      "step": 4025
    },
    {
      "epoch": 1.79,
      "learning_rate": 2.014814814814815e-05,
      "loss": 0.2531,
      "step": 4030
    },
    {
      "epoch": 1.8,
      "learning_rate": 2.0074074074074075e-05,
      "loss": 0.2241,
      "step": 4040
    },
    {
      "epoch": 1.8,
      "learning_rate": 2e-05,
      "loss": 0.2111,
      "step": 4050
    },
    {
      "epoch": 1.8,
      "eval_loss": 0.25542324781417847,
      "eval_runtime": 7.2106,
      "eval_samples_per_second": 138.823,
      "eval_steps_per_second": 34.81,
      "step": 4050
    },
    {
      "epoch": 1.8,
      "learning_rate": 1.9925925925925925e-05,
      "loss": 0.2535,
      "step": 4060
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.985185185185185e-05,
      "loss": 0.2385,
      "step": 4070
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.2562381625175476,
      "eval_runtime": 7.1846,
      "eval_samples_per_second": 139.325,
      "eval_steps_per_second": 34.936,
      "step": 4075
    },
    {
      "epoch": 1.81,
      "learning_rate": 1.9777777777777778e-05,
      "loss": 0.212,
      "step": 4080
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.9703703703703704e-05,
      "loss": 0.2094,
      "step": 4090
    },
    {
      "epoch": 1.82,
      "learning_rate": 1.962962962962963e-05,
      "loss": 0.2499,
      "step": 4100
    },
    {
      "epoch": 1.82,
      "eval_loss": 0.2557750344276428,
      "eval_runtime": 7.2707,
      "eval_samples_per_second": 137.676,
      "eval_steps_per_second": 34.522,
      "step": 4100
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.9555555555555557e-05,
      "loss": 0.2507,
      "step": 4110
    },
    {
      "epoch": 1.83,
      "learning_rate": 1.948148148148148e-05,
      "loss": 0.233,
      "step": 4120
    },
    {
      "epoch": 1.83,
      "eval_loss": 0.25592681765556335,
      "eval_runtime": 7.3242,
      "eval_samples_per_second": 136.671,
      "eval_steps_per_second": 34.27,
      "step": 4125
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.9407407407407407e-05,
      "loss": 0.2772,
      "step": 4130
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 0.2487,
      "step": 4140
    },
    {
      "epoch": 1.84,
      "learning_rate": 1.925925925925926e-05,
      "loss": 0.2538,
      "step": 4150
    },
    {
      "epoch": 1.84,
      "eval_loss": 0.25490960478782654,
      "eval_runtime": 7.3569,
      "eval_samples_per_second": 136.062,
      "eval_steps_per_second": 34.117,
      "step": 4150
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.9185185185185186e-05,
      "loss": 0.2681,
      "step": 4160
    },
    {
      "epoch": 1.85,
      "learning_rate": 1.9111111111111113e-05,
      "loss": 0.3042,
      "step": 4170
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.2556719183921814,
      "eval_runtime": 7.4839,
      "eval_samples_per_second": 133.753,
      "eval_steps_per_second": 33.539,
      "step": 4175
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.903703703703704e-05,
      "loss": 0.2017,
      "step": 4180
    },
    {
      "epoch": 1.86,
      "learning_rate": 1.8962962962962963e-05,
      "loss": 0.2162,
      "step": 4190
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.1989,
      "step": 4200
    },
    {
      "epoch": 1.87,
      "eval_loss": 0.25471413135528564,
      "eval_runtime": 7.4955,
      "eval_samples_per_second": 133.547,
      "eval_steps_per_second": 33.487,
      "step": 4200
    },
    {
      "epoch": 1.87,
      "learning_rate": 1.8814814814814816e-05,
      "loss": 0.2227,
      "step": 4210
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.8740740740740742e-05,
      "loss": 0.198,
      "step": 4220
    },
    {
      "epoch": 1.88,
      "eval_loss": 0.2542085349559784,
      "eval_runtime": 7.5713,
      "eval_samples_per_second": 132.211,
      "eval_steps_per_second": 33.152,
      "step": 4225
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.2839,
      "step": 4230
    },
    {
      "epoch": 1.88,
      "learning_rate": 1.8592592592592595e-05,
      "loss": 0.2969,
      "step": 4240
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.8518518518518518e-05,
      "loss": 0.2634,
      "step": 4250
    },
    {
      "epoch": 1.89,
      "eval_loss": 0.25297510623931885,
      "eval_runtime": 7.6201,
      "eval_samples_per_second": 131.363,
      "eval_steps_per_second": 32.939,
      "step": 4250
    },
    {
      "epoch": 1.89,
      "learning_rate": 1.8444444444444445e-05,
      "loss": 0.2182,
      "step": 4260
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.837037037037037e-05,
      "loss": 0.2684,
      "step": 4270
    },
    {
      "epoch": 1.9,
      "eval_loss": 0.2529478371143341,
      "eval_runtime": 7.6013,
      "eval_samples_per_second": 131.689,
      "eval_steps_per_second": 33.021,
      "step": 4275
    },
    {
      "epoch": 1.9,
      "learning_rate": 1.8296296296296298e-05,
      "loss": 0.302,
      "step": 4280
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.8222222222222224e-05,
      "loss": 0.195,
      "step": 4290
    },
    {
      "epoch": 1.91,
      "learning_rate": 1.814814814814815e-05,
      "loss": 0.1844,
      "step": 4300
    },
    {
      "epoch": 1.91,
      "eval_loss": 0.2522367835044861,
      "eval_runtime": 7.3231,
      "eval_samples_per_second": 136.692,
      "eval_steps_per_second": 34.275,
      "step": 4300
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.8074074074074074e-05,
      "loss": 0.2901,
      "step": 4310
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.8e-05,
      "loss": 0.2303,
      "step": 4320
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.252410352230072,
      "eval_runtime": 7.3587,
      "eval_samples_per_second": 136.029,
      "eval_steps_per_second": 34.109,
      "step": 4325
    },
    {
      "epoch": 1.92,
      "learning_rate": 1.7925925925925927e-05,
      "loss": 0.3308,
      "step": 4330
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.7851851851851853e-05,
      "loss": 0.2441,
      "step": 4340
    },
    {
      "epoch": 1.93,
      "learning_rate": 1.777777777777778e-05,
      "loss": 0.2126,
      "step": 4350
    },
    {
      "epoch": 1.93,
      "eval_loss": 0.25146955251693726,
      "eval_runtime": 7.3139,
      "eval_samples_per_second": 136.863,
      "eval_steps_per_second": 34.318,
      "step": 4350
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.7703703703703706e-05,
      "loss": 0.2338,
      "step": 4360
    },
    {
      "epoch": 1.94,
      "learning_rate": 1.762962962962963e-05,
      "loss": 0.2278,
      "step": 4370
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.2519494593143463,
      "eval_runtime": 7.2536,
      "eval_samples_per_second": 138.0,
      "eval_steps_per_second": 34.604,
      "step": 4375
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.7555555555555556e-05,
      "loss": 0.2033,
      "step": 4380
    },
    {
      "epoch": 1.95,
      "learning_rate": 1.7481481481481483e-05,
      "loss": 0.2524,
      "step": 4390
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.740740740740741e-05,
      "loss": 0.2692,
      "step": 4400
    },
    {
      "epoch": 1.96,
      "eval_loss": 0.25108450651168823,
      "eval_runtime": 7.1825,
      "eval_samples_per_second": 139.366,
      "eval_steps_per_second": 34.946,
      "step": 4400
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.2222,
      "step": 4410
    },
    {
      "epoch": 1.96,
      "learning_rate": 1.7259259259259262e-05,
      "loss": 0.2398,
      "step": 4420
    },
    {
      "epoch": 1.97,
      "eval_loss": 0.25103849172592163,
      "eval_runtime": 7.1987,
      "eval_samples_per_second": 139.053,
      "eval_steps_per_second": 34.867,
      "step": 4425
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7185185185185185e-05,
      "loss": 0.2295,
      "step": 4430
    },
    {
      "epoch": 1.97,
      "learning_rate": 1.7111111111111112e-05,
      "loss": 0.2153,
      "step": 4440
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.7037037037037038e-05,
      "loss": 0.2799,
      "step": 4450
    },
    {
      "epoch": 1.98,
      "eval_loss": 0.25079530477523804,
      "eval_runtime": 7.4048,
      "eval_samples_per_second": 135.183,
      "eval_steps_per_second": 33.897,
      "step": 4450
    },
    {
      "epoch": 1.98,
      "learning_rate": 1.6962962962962965e-05,
      "loss": 0.2143,
      "step": 4460
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.688888888888889e-05,
      "loss": 0.2492,
      "step": 4470
    },
    {
      "epoch": 1.99,
      "eval_loss": 0.25007081031799316,
      "eval_runtime": 7.3873,
      "eval_samples_per_second": 135.504,
      "eval_steps_per_second": 33.977,
      "step": 4475
    },
    {
      "epoch": 1.99,
      "learning_rate": 1.6814814814814818e-05,
      "loss": 0.2657,
      "step": 4480
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.674074074074074e-05,
      "loss": 0.2207,
      "step": 4490
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2202,
      "step": 4500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.24966831505298615,
      "eval_runtime": 7.3428,
      "eval_samples_per_second": 136.324,
      "eval_steps_per_second": 34.183,
      "step": 4500
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.6592592592592594e-05,
      "loss": 0.2467,
      "step": 4510
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.651851851851852e-05,
      "loss": 0.1736,
      "step": 4520
    },
    {
      "epoch": 2.01,
      "eval_loss": 0.25051918625831604,
      "eval_runtime": 7.2975,
      "eval_samples_per_second": 137.171,
      "eval_steps_per_second": 34.395,
      "step": 4525
    },
    {
      "epoch": 2.01,
      "learning_rate": 1.6444444444444447e-05,
      "loss": 0.1836,
      "step": 4530
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.6370370370370374e-05,
      "loss": 0.1811,
      "step": 4540
    },
    {
      "epoch": 2.02,
      "learning_rate": 1.62962962962963e-05,
      "loss": 0.204,
      "step": 4550
    },
    {
      "epoch": 2.02,
      "eval_loss": 0.25177687406539917,
      "eval_runtime": 7.1922,
      "eval_samples_per_second": 139.179,
      "eval_steps_per_second": 34.899,
      "step": 4550
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.6222222222222223e-05,
      "loss": 0.22,
      "step": 4560
    },
    {
      "epoch": 2.03,
      "learning_rate": 1.614814814814815e-05,
      "loss": 0.1853,
      "step": 4570
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.25192761421203613,
      "eval_runtime": 7.1908,
      "eval_samples_per_second": 139.205,
      "eval_steps_per_second": 34.906,
      "step": 4575
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.6074074074074076e-05,
      "loss": 0.1916,
      "step": 4580
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.2385,
      "step": 4590
    },
    {
      "epoch": 2.04,
      "learning_rate": 1.5925925925925926e-05,
      "loss": 0.1834,
      "step": 4600
    },
    {
      "epoch": 2.04,
      "eval_loss": 0.25096383690834045,
      "eval_runtime": 7.214,
      "eval_samples_per_second": 138.757,
      "eval_steps_per_second": 34.793,
      "step": 4600
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.5851851851851852e-05,
      "loss": 0.3046,
      "step": 4610
    },
    {
      "epoch": 2.05,
      "learning_rate": 1.577777777777778e-05,
      "loss": 0.2285,
      "step": 4620
    },
    {
      "epoch": 2.06,
      "eval_loss": 0.251299113035202,
      "eval_runtime": 7.2235,
      "eval_samples_per_second": 138.575,
      "eval_steps_per_second": 34.748,
      "step": 4625
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.5703703703703705e-05,
      "loss": 0.213,
      "step": 4630
    },
    {
      "epoch": 2.06,
      "learning_rate": 1.5629629629629632e-05,
      "loss": 0.1222,
      "step": 4640
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.179,
      "step": 4650
    },
    {
      "epoch": 2.07,
      "eval_loss": 0.25189557671546936,
      "eval_runtime": 7.2744,
      "eval_samples_per_second": 137.607,
      "eval_steps_per_second": 34.505,
      "step": 4650
    },
    {
      "epoch": 2.07,
      "learning_rate": 1.548148148148148e-05,
      "loss": 0.1749,
      "step": 4660
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5407407407407408e-05,
      "loss": 0.2163,
      "step": 4670
    },
    {
      "epoch": 2.08,
      "eval_loss": 0.25133654475212097,
      "eval_runtime": 7.2066,
      "eval_samples_per_second": 138.901,
      "eval_steps_per_second": 34.829,
      "step": 4675
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 0.1791,
      "step": 4680
    },
    {
      "epoch": 2.08,
      "learning_rate": 1.5259259259259258e-05,
      "loss": 0.1868,
      "step": 4690
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5185185185185186e-05,
      "loss": 0.2034,
      "step": 4700
    },
    {
      "epoch": 2.09,
      "eval_loss": 0.251429945230484,
      "eval_runtime": 7.2815,
      "eval_samples_per_second": 137.471,
      "eval_steps_per_second": 34.471,
      "step": 4700
    },
    {
      "epoch": 2.09,
      "learning_rate": 1.5111111111111112e-05,
      "loss": 0.1877,
      "step": 4710
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.5037037037037039e-05,
      "loss": 0.2275,
      "step": 4720
    },
    {
      "epoch": 2.1,
      "eval_loss": 0.25081896781921387,
      "eval_runtime": 7.1988,
      "eval_samples_per_second": 139.051,
      "eval_steps_per_second": 34.867,
      "step": 4725
    },
    {
      "epoch": 2.1,
      "learning_rate": 1.4962962962962965e-05,
      "loss": 0.2512,
      "step": 4730
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4888888888888888e-05,
      "loss": 0.1883,
      "step": 4740
    },
    {
      "epoch": 2.11,
      "learning_rate": 1.4814814814814815e-05,
      "loss": 0.2328,
      "step": 4750
    },
    {
      "epoch": 2.11,
      "eval_loss": 0.2509821653366089,
      "eval_runtime": 9.4138,
      "eval_samples_per_second": 106.334,
      "eval_steps_per_second": 26.663,
      "step": 4750
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4740740740740741e-05,
      "loss": 0.2146,
      "step": 4760
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4666666666666668e-05,
      "loss": 0.1935,
      "step": 4770
    },
    {
      "epoch": 2.12,
      "eval_loss": 0.2518753111362457,
      "eval_runtime": 7.4318,
      "eval_samples_per_second": 134.691,
      "eval_steps_per_second": 33.774,
      "step": 4775
    },
    {
      "epoch": 2.12,
      "learning_rate": 1.4592592592592594e-05,
      "loss": 0.2174,
      "step": 4780
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.4518518518518521e-05,
      "loss": 0.1887,
      "step": 4790
    },
    {
      "epoch": 2.13,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 0.1613,
      "step": 4800
    },
    {
      "epoch": 2.13,
      "eval_loss": 0.251521497964859,
      "eval_runtime": 7.26,
      "eval_samples_per_second": 137.88,
      "eval_steps_per_second": 34.573,
      "step": 4800
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.437037037037037e-05,
      "loss": 0.2027,
      "step": 4810
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.4296296296296297e-05,
      "loss": 0.2451,
      "step": 4820
    },
    {
      "epoch": 2.14,
      "eval_loss": 0.2503267526626587,
      "eval_runtime": 7.228,
      "eval_samples_per_second": 138.488,
      "eval_steps_per_second": 34.726,
      "step": 4825
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4222222222222224e-05,
      "loss": 0.199,
      "step": 4830
    },
    {
      "epoch": 2.15,
      "learning_rate": 1.4148148148148148e-05,
      "loss": 0.1797,
      "step": 4840
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.4074074074074075e-05,
      "loss": 0.199,
      "step": 4850
    },
    {
      "epoch": 2.16,
      "eval_loss": 0.2507803440093994,
      "eval_runtime": 7.2959,
      "eval_samples_per_second": 137.2,
      "eval_steps_per_second": 34.403,
      "step": 4850
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 0.2138,
      "step": 4860
    },
    {
      "epoch": 2.16,
      "learning_rate": 1.3925925925925926e-05,
      "loss": 0.2486,
      "step": 4870
    },
    {
      "epoch": 2.17,
      "eval_loss": 0.25041449069976807,
      "eval_runtime": 7.2127,
      "eval_samples_per_second": 138.783,
      "eval_steps_per_second": 34.8,
      "step": 4875
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3851851851851853e-05,
      "loss": 0.1906,
      "step": 4880
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3777777777777778e-05,
      "loss": 0.2375,
      "step": 4890
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.3703703703703704e-05,
      "loss": 0.2124,
      "step": 4900
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.2496342658996582,
      "eval_runtime": 7.2877,
      "eval_samples_per_second": 137.354,
      "eval_steps_per_second": 34.441,
      "step": 4900
    },
    {
      "epoch": 2.18,
      "learning_rate": 1.362962962962963e-05,
      "loss": 0.1827,
      "step": 4910
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.3555555555555557e-05,
      "loss": 0.2032,
      "step": 4920
    },
    {
      "epoch": 2.19,
      "eval_loss": 0.2495114803314209,
      "eval_runtime": 7.1954,
      "eval_samples_per_second": 139.117,
      "eval_steps_per_second": 34.883,
      "step": 4925
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.348148148148148e-05,
      "loss": 0.133,
      "step": 4930
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3407407407407407e-05,
      "loss": 0.2068,
      "step": 4940
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.1772,
      "step": 4950
    },
    {
      "epoch": 2.2,
      "eval_loss": 0.25019142031669617,
      "eval_runtime": 7.291,
      "eval_samples_per_second": 137.293,
      "eval_steps_per_second": 34.426,
      "step": 4950
    },
    {
      "epoch": 2.2,
      "learning_rate": 1.325925925925926e-05,
      "loss": 0.213,
      "step": 4960
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3185185185185186e-05,
      "loss": 0.1879,
      "step": 4970
    },
    {
      "epoch": 2.21,
      "eval_loss": 0.24890127778053284,
      "eval_runtime": 7.2295,
      "eval_samples_per_second": 138.46,
      "eval_steps_per_second": 34.719,
      "step": 4975
    },
    {
      "epoch": 2.21,
      "learning_rate": 1.3111111111111113e-05,
      "loss": 0.1829,
      "step": 4980
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.3037037037037036e-05,
      "loss": 0.211,
      "step": 4990
    },
    {
      "epoch": 2.22,
      "learning_rate": 1.2962962962962962e-05,
      "loss": 0.1554,
      "step": 5000
    },
    {
      "epoch": 2.22,
      "eval_loss": 0.24963602423667908,
      "eval_runtime": 7.2892,
      "eval_samples_per_second": 137.326,
      "eval_steps_per_second": 34.434,
      "step": 5000
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.2888888888888889e-05,
      "loss": 0.1922,
      "step": 5010
    },
    {
      "epoch": 2.23,
      "learning_rate": 1.2814814814814815e-05,
      "loss": 0.2178,
      "step": 5020
    },
    {
      "epoch": 2.23,
      "eval_loss": 0.24953505396842957,
      "eval_runtime": 7.1973,
      "eval_samples_per_second": 139.08,
      "eval_steps_per_second": 34.874,
      "step": 5025
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2740740740740742e-05,
      "loss": 0.1909,
      "step": 5030
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2666666666666668e-05,
      "loss": 0.2111,
      "step": 5040
    },
    {
      "epoch": 2.24,
      "learning_rate": 1.2592592592592592e-05,
      "loss": 0.1936,
      "step": 5050
    },
    {
      "epoch": 2.24,
      "eval_loss": 0.249254047870636,
      "eval_runtime": 7.2689,
      "eval_samples_per_second": 137.711,
      "eval_steps_per_second": 34.531,
      "step": 5050
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.2518518518518518e-05,
      "loss": 0.1913,
      "step": 5060
    },
    {
      "epoch": 2.25,
      "learning_rate": 1.2444444444444445e-05,
      "loss": 0.2,
      "step": 5070
    },
    {
      "epoch": 2.26,
      "eval_loss": 0.2488778978586197,
      "eval_runtime": 7.2184,
      "eval_samples_per_second": 138.674,
      "eval_steps_per_second": 34.772,
      "step": 5075
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.2370370370370371e-05,
      "loss": 0.1723,
      "step": 5080
    },
    {
      "epoch": 2.26,
      "learning_rate": 1.2296296296296298e-05,
      "loss": 0.1875,
      "step": 5090
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 0.185,
      "step": 5100
    },
    {
      "epoch": 2.27,
      "eval_loss": 0.24892283976078033,
      "eval_runtime": 7.3812,
      "eval_samples_per_second": 135.615,
      "eval_steps_per_second": 34.005,
      "step": 5100
    },
    {
      "epoch": 2.27,
      "learning_rate": 1.2148148148148149e-05,
      "loss": 0.2168,
      "step": 5110
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.2074074074074075e-05,
      "loss": 0.1783,
      "step": 5120
    },
    {
      "epoch": 2.28,
      "eval_loss": 0.2482217699289322,
      "eval_runtime": 7.3518,
      "eval_samples_per_second": 136.157,
      "eval_steps_per_second": 34.141,
      "step": 5125
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.2e-05,
      "loss": 0.1643,
      "step": 5130
    },
    {
      "epoch": 2.28,
      "learning_rate": 1.1925925925925927e-05,
      "loss": 0.2211,
      "step": 5140
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1851851851851853e-05,
      "loss": 0.2276,
      "step": 5150
    },
    {
      "epoch": 2.29,
      "eval_loss": 0.2477390617132187,
      "eval_runtime": 7.5029,
      "eval_samples_per_second": 133.414,
      "eval_steps_per_second": 33.454,
      "step": 5150
    },
    {
      "epoch": 2.29,
      "learning_rate": 1.1777777777777778e-05,
      "loss": 0.1951,
      "step": 5160
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1703703703703705e-05,
      "loss": 0.2134,
      "step": 5170
    },
    {
      "epoch": 2.3,
      "eval_loss": 0.24737687408924103,
      "eval_runtime": 7.4753,
      "eval_samples_per_second": 133.907,
      "eval_steps_per_second": 33.577,
      "step": 5175
    },
    {
      "epoch": 2.3,
      "learning_rate": 1.1629629629629631e-05,
      "loss": 0.2208,
      "step": 5180
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.1555555555555556e-05,
      "loss": 0.1714,
      "step": 5190
    },
    {
      "epoch": 2.31,
      "learning_rate": 1.1481481481481482e-05,
      "loss": 0.1747,
      "step": 5200
    },
    {
      "epoch": 2.31,
      "eval_loss": 0.2470199465751648,
      "eval_runtime": 7.6413,
      "eval_samples_per_second": 130.999,
      "eval_steps_per_second": 32.848,
      "step": 5200
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1407407407407409e-05,
      "loss": 0.209,
      "step": 5210
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 0.2121,
      "step": 5220
    },
    {
      "epoch": 2.32,
      "eval_loss": 0.2458607703447342,
      "eval_runtime": 7.515,
      "eval_samples_per_second": 133.2,
      "eval_steps_per_second": 33.4,
      "step": 5225
    },
    {
      "epoch": 2.32,
      "learning_rate": 1.125925925925926e-05,
      "loss": 0.2266,
      "step": 5230
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.1185185185185187e-05,
      "loss": 0.1754,
      "step": 5240
    },
    {
      "epoch": 2.33,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.2199,
      "step": 5250
    },
    {
      "epoch": 2.33,
      "eval_loss": 0.2461249679327011,
      "eval_runtime": 7.5406,
      "eval_samples_per_second": 132.748,
      "eval_steps_per_second": 33.286,
      "step": 5250
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.1037037037037038e-05,
      "loss": 0.2609,
      "step": 5260
    },
    {
      "epoch": 2.34,
      "learning_rate": 1.0962962962962963e-05,
      "loss": 0.2708,
      "step": 5270
    },
    {
      "epoch": 2.34,
      "eval_loss": 0.24557095766067505,
      "eval_runtime": 7.3992,
      "eval_samples_per_second": 135.286,
      "eval_steps_per_second": 33.923,
      "step": 5275
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.088888888888889e-05,
      "loss": 0.1986,
      "step": 5280
    },
    {
      "epoch": 2.35,
      "learning_rate": 1.0814814814814814e-05,
      "loss": 0.1937,
      "step": 5290
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.074074074074074e-05,
      "loss": 0.2252,
      "step": 5300
    },
    {
      "epoch": 2.36,
      "eval_loss": 0.24577473104000092,
      "eval_runtime": 7.388,
      "eval_samples_per_second": 135.49,
      "eval_steps_per_second": 33.974,
      "step": 5300
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.1642,
      "step": 5310
    },
    {
      "epoch": 2.36,
      "learning_rate": 1.0592592592592592e-05,
      "loss": 0.1921,
      "step": 5320
    },
    {
      "epoch": 2.37,
      "eval_loss": 0.24629971385002136,
      "eval_runtime": 7.2504,
      "eval_samples_per_second": 138.061,
      "eval_steps_per_second": 34.619,
      "step": 5325
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.0518518518518519e-05,
      "loss": 0.1795,
      "step": 5330
    },
    {
      "epoch": 2.37,
      "learning_rate": 1.0444444444444445e-05,
      "loss": 0.2051,
      "step": 5340
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.037037037037037e-05,
      "loss": 0.1627,
      "step": 5350
    },
    {
      "epoch": 2.38,
      "eval_loss": 0.24658331274986267,
      "eval_runtime": 7.6185,
      "eval_samples_per_second": 131.39,
      "eval_steps_per_second": 32.946,
      "step": 5350
    },
    {
      "epoch": 2.38,
      "learning_rate": 1.0296296296296296e-05,
      "loss": 0.1746,
      "step": 5360
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0222222222222223e-05,
      "loss": 0.1988,
      "step": 5370
    },
    {
      "epoch": 2.39,
      "eval_loss": 0.24601979553699493,
      "eval_runtime": 7.5137,
      "eval_samples_per_second": 133.223,
      "eval_steps_per_second": 33.406,
      "step": 5375
    },
    {
      "epoch": 2.39,
      "learning_rate": 1.0148148148148148e-05,
      "loss": 0.1994,
      "step": 5380
    },
    {
      "epoch": 2.4,
      "learning_rate": 1.0074074074074074e-05,
      "loss": 0.191,
      "step": 5390
    },
    {
      "epoch": 2.4,
      "learning_rate": 1e-05,
      "loss": 0.2308,
      "step": 5400
    },
    {
      "epoch": 2.4,
      "eval_loss": 0.24538284540176392,
      "eval_runtime": 7.6411,
      "eval_samples_per_second": 131.002,
      "eval_steps_per_second": 32.849,
      "step": 5400
    },
    {
      "epoch": 2.4,
      "learning_rate": 9.925925925925926e-06,
      "loss": 0.2331,
      "step": 5410
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.851851851851852e-06,
      "loss": 0.138,
      "step": 5420
    },
    {
      "epoch": 2.41,
      "eval_loss": 0.24550172686576843,
      "eval_runtime": 7.6397,
      "eval_samples_per_second": 131.026,
      "eval_steps_per_second": 32.855,
      "step": 5425
    },
    {
      "epoch": 2.41,
      "learning_rate": 9.777777777777779e-06,
      "loss": 0.2266,
      "step": 5430
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.703703703703703e-06,
      "loss": 0.1693,
      "step": 5440
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.62962962962963e-06,
      "loss": 0.1745,
      "step": 5450
    },
    {
      "epoch": 2.42,
      "eval_loss": 0.2455550730228424,
      "eval_runtime": 7.6397,
      "eval_samples_per_second": 131.025,
      "eval_steps_per_second": 32.855,
      "step": 5450
    },
    {
      "epoch": 2.43,
      "learning_rate": 9.555555555555556e-06,
      "loss": 0.1472,
      "step": 5460
    },
    {
      "epoch": 2.43,
      "learning_rate": 9.481481481481481e-06,
      "loss": 0.1712,
      "step": 5470
    },
    {
      "epoch": 2.43,
      "eval_loss": 0.24605458974838257,
      "eval_runtime": 8.1344,
      "eval_samples_per_second": 123.058,
      "eval_steps_per_second": 30.857,
      "step": 5475
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.407407407407408e-06,
      "loss": 0.1755,
      "step": 5480
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.1984,
      "step": 5490
    },
    {
      "epoch": 2.44,
      "learning_rate": 9.259259259259259e-06,
      "loss": 0.1763,
      "step": 5500
    },
    {
      "epoch": 2.44,
      "eval_loss": 0.24583406746387482,
      "eval_runtime": 10.3194,
      "eval_samples_per_second": 97.002,
      "eval_steps_per_second": 24.323,
      "step": 5500
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.185185185185186e-06,
      "loss": 0.2261,
      "step": 5510
    },
    {
      "epoch": 2.45,
      "learning_rate": 9.111111111111112e-06,
      "loss": 0.2068,
      "step": 5520
    },
    {
      "epoch": 2.46,
      "eval_loss": 0.24605007469654083,
      "eval_runtime": 7.6128,
      "eval_samples_per_second": 131.489,
      "eval_steps_per_second": 32.971,
      "step": 5525
    },
    {
      "epoch": 2.46,
      "learning_rate": 9.037037037037037e-06,
      "loss": 0.2097,
      "step": 5530
    },
    {
      "epoch": 2.46,
      "learning_rate": 8.962962962962963e-06,
      "loss": 0.1733,
      "step": 5540
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.88888888888889e-06,
      "loss": 0.2097,
      "step": 5550
    },
    {
      "epoch": 2.47,
      "eval_loss": 0.24533824622631073,
      "eval_runtime": 7.6087,
      "eval_samples_per_second": 131.559,
      "eval_steps_per_second": 32.988,
      "step": 5550
    },
    {
      "epoch": 2.47,
      "learning_rate": 8.814814814814815e-06,
      "loss": 0.1896,
      "step": 5560
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.740740740740741e-06,
      "loss": 0.1945,
      "step": 5570
    },
    {
      "epoch": 2.48,
      "eval_loss": 0.24509920179843903,
      "eval_runtime": 7.6048,
      "eval_samples_per_second": 131.627,
      "eval_steps_per_second": 33.005,
      "step": 5575
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.2036,
      "step": 5580
    },
    {
      "epoch": 2.48,
      "learning_rate": 8.592592592592593e-06,
      "loss": 0.1933,
      "step": 5590
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.518518518518519e-06,
      "loss": 0.1906,
      "step": 5600
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.24449166655540466,
      "eval_runtime": 7.5807,
      "eval_samples_per_second": 132.047,
      "eval_steps_per_second": 33.111,
      "step": 5600
    },
    {
      "epoch": 2.49,
      "learning_rate": 8.444444444444446e-06,
      "loss": 0.1655,
      "step": 5610
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.37037037037037e-06,
      "loss": 0.1681,
      "step": 5620
    },
    {
      "epoch": 2.5,
      "eval_loss": 0.24523867666721344,
      "eval_runtime": 7.5355,
      "eval_samples_per_second": 132.839,
      "eval_steps_per_second": 33.309,
      "step": 5625
    },
    {
      "epoch": 2.5,
      "learning_rate": 8.296296296296297e-06,
      "loss": 0.202,
      "step": 5630
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.222222222222223e-06,
      "loss": 0.1795,
      "step": 5640
    },
    {
      "epoch": 2.51,
      "learning_rate": 8.14814814814815e-06,
      "loss": 0.1732,
      "step": 5650
    },
    {
      "epoch": 2.51,
      "eval_loss": 0.24490933120250702,
      "eval_runtime": 7.4917,
      "eval_samples_per_second": 133.614,
      "eval_steps_per_second": 33.504,
      "step": 5650
    },
    {
      "epoch": 2.52,
      "learning_rate": 8.074074074074075e-06,
      "loss": 0.1765,
      "step": 5660
    },
    {
      "epoch": 2.52,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.2129,
      "step": 5670
    },
    {
      "epoch": 2.52,
      "eval_loss": 0.2450021654367447,
      "eval_runtime": 7.5776,
      "eval_samples_per_second": 132.099,
      "eval_steps_per_second": 33.124,
      "step": 5675
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.925925925925926e-06,
      "loss": 0.239,
      "step": 5680
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.851851851851853e-06,
      "loss": 0.153,
      "step": 5690
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.777777777777777e-06,
      "loss": 0.1749,
      "step": 5700
    },
    {
      "epoch": 2.53,
      "eval_loss": 0.244710773229599,
      "eval_runtime": 7.2355,
      "eval_samples_per_second": 138.346,
      "eval_steps_per_second": 34.69,
      "step": 5700
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.703703703703704e-06,
      "loss": 0.2436,
      "step": 5710
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.629629629629629e-06,
      "loss": 0.217,
      "step": 5720
    },
    {
      "epoch": 2.54,
      "eval_loss": 0.24449826776981354,
      "eval_runtime": 7.2691,
      "eval_samples_per_second": 137.706,
      "eval_steps_per_second": 34.53,
      "step": 5725
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.555555555555556e-06,
      "loss": 0.1999,
      "step": 5730
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.481481481481483e-06,
      "loss": 0.1981,
      "step": 5740
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.4074074074074075e-06,
      "loss": 0.2359,
      "step": 5750
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.24463185667991638,
      "eval_runtime": 7.1779,
      "eval_samples_per_second": 139.456,
      "eval_steps_per_second": 34.968,
      "step": 5750
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.333333333333334e-06,
      "loss": 0.1824,
      "step": 5760
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.2592592592592605e-06,
      "loss": 0.1813,
      "step": 5770
    },
    {
      "epoch": 2.57,
      "eval_loss": 0.24394886195659637,
      "eval_runtime": 7.21,
      "eval_samples_per_second": 138.835,
      "eval_steps_per_second": 34.813,
      "step": 5775
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.185185185185185e-06,
      "loss": 0.2108,
      "step": 5780
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.111111111111112e-06,
      "loss": 0.2403,
      "step": 5790
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.0370370370370375e-06,
      "loss": 0.1771,
      "step": 5800
    },
    {
      "epoch": 2.58,
      "eval_loss": 0.2437148541212082,
      "eval_runtime": 7.3132,
      "eval_samples_per_second": 136.876,
      "eval_steps_per_second": 34.322,
      "step": 5800
    },
    {
      "epoch": 2.58,
      "learning_rate": 6.962962962962963e-06,
      "loss": 0.2451,
      "step": 5810
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.888888888888889e-06,
      "loss": 0.2009,
      "step": 5820
    },
    {
      "epoch": 2.59,
      "eval_loss": 0.2434249371290207,
      "eval_runtime": 7.2848,
      "eval_samples_per_second": 137.41,
      "eval_steps_per_second": 34.455,
      "step": 5825
    },
    {
      "epoch": 2.59,
      "learning_rate": 6.814814814814815e-06,
      "loss": 0.1734,
      "step": 5830
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.74074074074074e-06,
      "loss": 0.2106,
      "step": 5840
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2255,
      "step": 5850
    },
    {
      "epoch": 2.6,
      "eval_loss": 0.24379944801330566,
      "eval_runtime": 7.3588,
      "eval_samples_per_second": 136.028,
      "eval_steps_per_second": 34.109,
      "step": 5850
    },
    {
      "epoch": 2.6,
      "learning_rate": 6.592592592592593e-06,
      "loss": 0.2042,
      "step": 5860
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.518518518518518e-06,
      "loss": 0.1645,
      "step": 5870
    },
    {
      "epoch": 2.61,
      "eval_loss": 0.24349051713943481,
      "eval_runtime": 7.3319,
      "eval_samples_per_second": 136.527,
      "eval_steps_per_second": 34.234,
      "step": 5875
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.4444444444444445e-06,
      "loss": 0.1971,
      "step": 5880
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.370370370370371e-06,
      "loss": 0.1939,
      "step": 5890
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.296296296296296e-06,
      "loss": 0.2087,
      "step": 5900
    },
    {
      "epoch": 2.62,
      "eval_loss": 0.24342414736747742,
      "eval_runtime": 7.3018,
      "eval_samples_per_second": 137.09,
      "eval_steps_per_second": 34.375,
      "step": 5900
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.222222222222222e-06,
      "loss": 0.2373,
      "step": 5910
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.148148148148149e-06,
      "loss": 0.2125,
      "step": 5920
    },
    {
      "epoch": 2.63,
      "eval_loss": 0.24332143366336823,
      "eval_runtime": 7.221,
      "eval_samples_per_second": 138.623,
      "eval_steps_per_second": 34.76,
      "step": 5925
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.0740740740740745e-06,
      "loss": 0.2188,
      "step": 5930
    },
    {
      "epoch": 2.64,
      "learning_rate": 6e-06,
      "loss": 0.1756,
      "step": 5940
    },
    {
      "epoch": 2.64,
      "learning_rate": 5.925925925925927e-06,
      "loss": 0.2484,
      "step": 5950
    },
    {
      "epoch": 2.64,
      "eval_loss": 0.24283772706985474,
      "eval_runtime": 7.2081,
      "eval_samples_per_second": 138.871,
      "eval_steps_per_second": 34.822,
      "step": 5950
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.851851851851852e-06,
      "loss": 0.201,
      "step": 5960
    },
    {
      "epoch": 2.65,
      "learning_rate": 5.777777777777778e-06,
      "loss": 0.1724,
      "step": 5970
    },
    {
      "epoch": 2.66,
      "eval_loss": 0.24253526329994202,
      "eval_runtime": 7.2139,
      "eval_samples_per_second": 138.76,
      "eval_steps_per_second": 34.794,
      "step": 5975
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.7037037037037045e-06,
      "loss": 0.2071,
      "step": 5980
    },
    {
      "epoch": 2.66,
      "learning_rate": 5.62962962962963e-06,
      "loss": 0.2026,
      "step": 5990
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.18,
      "step": 6000
    },
    {
      "epoch": 2.67,
      "eval_loss": 0.2427457869052887,
      "eval_runtime": 7.24,
      "eval_samples_per_second": 138.259,
      "eval_steps_per_second": 34.668,
      "step": 6000
    },
    {
      "epoch": 2.67,
      "learning_rate": 5.4814814814814815e-06,
      "loss": 0.2491,
      "step": 6010
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.407407407407407e-06,
      "loss": 0.1878,
      "step": 6020
    },
    {
      "epoch": 2.68,
      "eval_loss": 0.2427491545677185,
      "eval_runtime": 7.2356,
      "eval_samples_per_second": 138.343,
      "eval_steps_per_second": 34.689,
      "step": 6025
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.196,
      "step": 6030
    },
    {
      "epoch": 2.68,
      "learning_rate": 5.259259259259259e-06,
      "loss": 0.1861,
      "step": 6040
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.185185185185185e-06,
      "loss": 0.2131,
      "step": 6050
    },
    {
      "epoch": 2.69,
      "eval_loss": 0.24262972176074982,
      "eval_runtime": 7.2135,
      "eval_samples_per_second": 138.768,
      "eval_steps_per_second": 34.796,
      "step": 6050
    },
    {
      "epoch": 2.69,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 0.1904,
      "step": 6060
    },
    {
      "epoch": 2.7,
      "learning_rate": 5.037037037037037e-06,
      "loss": 0.1677,
      "step": 6070
    },
    {
      "epoch": 2.7,
      "eval_loss": 0.24258430302143097,
      "eval_runtime": 7.38,
      "eval_samples_per_second": 135.638,
      "eval_steps_per_second": 34.011,
      "step": 6075
    },
    {
      "epoch": 2.7,
      "learning_rate": 4.962962962962963e-06,
      "loss": 0.1728,
      "step": 6080
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.888888888888889e-06,
      "loss": 0.2153,
      "step": 6090
    },
    {
      "epoch": 2.71,
      "learning_rate": 4.814814814814815e-06,
      "loss": 0.2097,
      "step": 6100
    },
    {
      "epoch": 2.71,
      "eval_loss": 0.2426908016204834,
      "eval_runtime": 7.3472,
      "eval_samples_per_second": 136.243,
      "eval_steps_per_second": 34.163,
      "step": 6100
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.740740740740741e-06,
      "loss": 0.1584,
      "step": 6110
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.2014,
      "step": 6120
    },
    {
      "epoch": 2.72,
      "eval_loss": 0.24265140295028687,
      "eval_runtime": 7.3411,
      "eval_samples_per_second": 136.355,
      "eval_steps_per_second": 34.191,
      "step": 6125
    },
    {
      "epoch": 2.72,
      "learning_rate": 4.592592592592593e-06,
      "loss": 0.2504,
      "step": 6130
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.5185185185185185e-06,
      "loss": 0.207,
      "step": 6140
    },
    {
      "epoch": 2.73,
      "learning_rate": 4.444444444444445e-06,
      "loss": 0.1777,
      "step": 6150
    },
    {
      "epoch": 2.73,
      "eval_loss": 0.24221929907798767,
      "eval_runtime": 7.2165,
      "eval_samples_per_second": 138.709,
      "eval_steps_per_second": 34.781,
      "step": 6150
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.370370370370371e-06,
      "loss": 0.1982,
      "step": 6160
    },
    {
      "epoch": 2.74,
      "learning_rate": 4.296296296296296e-06,
      "loss": 0.189,
      "step": 6170
    },
    {
      "epoch": 2.74,
      "eval_loss": 0.24199625849723816,
      "eval_runtime": 7.2328,
      "eval_samples_per_second": 138.397,
      "eval_steps_per_second": 34.703,
      "step": 6175
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.222222222222223e-06,
      "loss": 0.1895,
      "step": 6180
    },
    {
      "epoch": 2.75,
      "learning_rate": 4.1481481481481485e-06,
      "loss": 0.2303,
      "step": 6190
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.074074074074075e-06,
      "loss": 0.2499,
      "step": 6200
    },
    {
      "epoch": 2.76,
      "eval_loss": 0.24183855950832367,
      "eval_runtime": 7.2222,
      "eval_samples_per_second": 138.6,
      "eval_steps_per_second": 34.754,
      "step": 6200
    },
    {
      "epoch": 2.76,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.2105,
      "step": 6210
    },
    {
      "epoch": 2.76,
      "learning_rate": 3.925925925925926e-06,
      "loss": 0.1827,
      "step": 6220
    },
    {
      "epoch": 2.77,
      "eval_loss": 0.24157124757766724,
      "eval_runtime": 7.2228,
      "eval_samples_per_second": 138.589,
      "eval_steps_per_second": 34.751,
      "step": 6225
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.851851851851852e-06,
      "loss": 0.1625,
      "step": 6230
    },
    {
      "epoch": 2.77,
      "learning_rate": 3.777777777777778e-06,
      "loss": 0.2046,
      "step": 6240
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.7037037037037037e-06,
      "loss": 0.1872,
      "step": 6250
    },
    {
      "epoch": 2.78,
      "eval_loss": 0.24152511358261108,
      "eval_runtime": 7.2189,
      "eval_samples_per_second": 138.665,
      "eval_steps_per_second": 34.77,
      "step": 6250
    },
    {
      "epoch": 2.78,
      "learning_rate": 3.6296296296296302e-06,
      "loss": 0.1476,
      "step": 6260
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.555555555555556e-06,
      "loss": 0.1948,
      "step": 6270
    },
    {
      "epoch": 2.79,
      "eval_loss": 0.24153129756450653,
      "eval_runtime": 7.2338,
      "eval_samples_per_second": 138.379,
      "eval_steps_per_second": 34.698,
      "step": 6275
    },
    {
      "epoch": 2.79,
      "learning_rate": 3.4814814814814816e-06,
      "loss": 0.184,
      "step": 6280
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.4074074074074077e-06,
      "loss": 0.2911,
      "step": 6290
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.2382,
      "step": 6300
    },
    {
      "epoch": 2.8,
      "eval_loss": 0.24131891131401062,
      "eval_runtime": 7.204,
      "eval_samples_per_second": 138.95,
      "eval_steps_per_second": 34.842,
      "step": 6300
    },
    {
      "epoch": 2.8,
      "learning_rate": 3.259259259259259e-06,
      "loss": 0.2274,
      "step": 6310
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.1851851851851855e-06,
      "loss": 0.2495,
      "step": 6320
    },
    {
      "epoch": 2.81,
      "eval_loss": 0.24129566550254822,
      "eval_runtime": 7.2128,
      "eval_samples_per_second": 138.78,
      "eval_steps_per_second": 34.799,
      "step": 6325
    },
    {
      "epoch": 2.81,
      "learning_rate": 3.111111111111111e-06,
      "loss": 0.1895,
      "step": 6330
    },
    {
      "epoch": 2.82,
      "learning_rate": 3.0370370370370372e-06,
      "loss": 0.2067,
      "step": 6340
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.9629629629629633e-06,
      "loss": 0.2258,
      "step": 6350
    },
    {
      "epoch": 2.82,
      "eval_loss": 0.24125269055366516,
      "eval_runtime": 7.2125,
      "eval_samples_per_second": 138.787,
      "eval_steps_per_second": 34.801,
      "step": 6350
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.888888888888889e-06,
      "loss": 0.1759,
      "step": 6360
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.814814814814815e-06,
      "loss": 0.1618,
      "step": 6370
    },
    {
      "epoch": 2.83,
      "eval_loss": 0.24155820906162262,
      "eval_runtime": 7.2268,
      "eval_samples_per_second": 138.511,
      "eval_steps_per_second": 34.732,
      "step": 6375
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.7407407407407407e-06,
      "loss": 0.1782,
      "step": 6380
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.666666666666667e-06,
      "loss": 0.1813,
      "step": 6390
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.5925925925925925e-06,
      "loss": 0.1605,
      "step": 6400
    },
    {
      "epoch": 2.84,
      "eval_loss": 0.24137118458747864,
      "eval_runtime": 7.2587,
      "eval_samples_per_second": 137.903,
      "eval_steps_per_second": 34.579,
      "step": 6400
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.5185185185185186e-06,
      "loss": 0.1751,
      "step": 6410
    },
    {
      "epoch": 2.85,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 0.1712,
      "step": 6420
    },
    {
      "epoch": 2.86,
      "eval_loss": 0.24129529297351837,
      "eval_runtime": 7.385,
      "eval_samples_per_second": 135.545,
      "eval_steps_per_second": 33.988,
      "step": 6425
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.3703703703703703e-06,
      "loss": 0.2067,
      "step": 6430
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.2962962962962964e-06,
      "loss": 0.177,
      "step": 6440
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 0.2021,
      "step": 6450
    },
    {
      "epoch": 2.87,
      "eval_loss": 0.24113218486309052,
      "eval_runtime": 7.1916,
      "eval_samples_per_second": 139.191,
      "eval_steps_per_second": 34.902,
      "step": 6450
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.148148148148148e-06,
      "loss": 0.1895,
      "step": 6460
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.0740740740740742e-06,
      "loss": 0.1985,
      "step": 6470
    },
    {
      "epoch": 2.88,
      "eval_loss": 0.24104492366313934,
      "eval_runtime": 7.2162,
      "eval_samples_per_second": 138.715,
      "eval_steps_per_second": 34.783,
      "step": 6475
    },
    {
      "epoch": 2.88,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.1742,
      "step": 6480
    },
    {
      "epoch": 2.88,
      "learning_rate": 1.925925925925926e-06,
      "loss": 0.2035,
      "step": 6490
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.8518518518518519e-06,
      "loss": 0.1773,
      "step": 6500
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.24099375307559967,
      "eval_runtime": 7.2337,
      "eval_samples_per_second": 138.38,
      "eval_steps_per_second": 34.699,
      "step": 6500
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.777777777777778e-06,
      "loss": 0.1872,
      "step": 6510
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.7037037037037038e-06,
      "loss": 0.2152,
      "step": 6520
    },
    {
      "epoch": 2.9,
      "eval_loss": 0.24097038805484772,
      "eval_runtime": 7.2148,
      "eval_samples_per_second": 138.743,
      "eval_steps_per_second": 34.79,
      "step": 6525
    },
    {
      "epoch": 2.9,
      "learning_rate": 1.6296296296296295e-06,
      "loss": 0.2588,
      "step": 6530
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.5555555555555556e-06,
      "loss": 0.2228,
      "step": 6540
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.4814814814814817e-06,
      "loss": 0.2225,
      "step": 6550
    },
    {
      "epoch": 2.91,
      "eval_loss": 0.24102109670639038,
      "eval_runtime": 7.2724,
      "eval_samples_per_second": 137.644,
      "eval_steps_per_second": 34.514,
      "step": 6550
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.4074074074074075e-06,
      "loss": 0.2303,
      "step": 6560
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 0.1393,
      "step": 6570
    },
    {
      "epoch": 2.92,
      "eval_loss": 0.24098366498947144,
      "eval_runtime": 7.2615,
      "eval_samples_per_second": 137.849,
      "eval_steps_per_second": 34.566,
      "step": 6575
    },
    {
      "epoch": 2.92,
      "learning_rate": 1.2592592592592593e-06,
      "loss": 0.1581,
      "step": 6580
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.1851851851851852e-06,
      "loss": 0.2093,
      "step": 6590
    },
    {
      "epoch": 2.93,
      "learning_rate": 1.1111111111111112e-06,
      "loss": 0.1631,
      "step": 6600
    },
    {
      "epoch": 2.93,
      "eval_loss": 0.24084864556789398,
      "eval_runtime": 7.2231,
      "eval_samples_per_second": 138.583,
      "eval_steps_per_second": 34.75,
      "step": 6600
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.0370370370370371e-06,
      "loss": 0.2028,
      "step": 6610
    },
    {
      "epoch": 2.94,
      "learning_rate": 9.62962962962963e-07,
      "loss": 0.2158,
      "step": 6620
    },
    {
      "epoch": 2.94,
      "eval_loss": 0.24080170691013336,
      "eval_runtime": 7.2315,
      "eval_samples_per_second": 138.423,
      "eval_steps_per_second": 34.709,
      "step": 6625
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.88888888888889e-07,
      "loss": 0.1775,
      "step": 6630
    },
    {
      "epoch": 2.95,
      "learning_rate": 8.148148148148147e-07,
      "loss": 0.1827,
      "step": 6640
    },
    {
      "epoch": 2.96,
      "learning_rate": 7.407407407407408e-07,
      "loss": 0.218,
      "step": 6650
    },
    {
      "epoch": 2.96,
      "eval_loss": 0.24078905582427979,
      "eval_runtime": 7.2318,
      "eval_samples_per_second": 138.417,
      "eval_steps_per_second": 34.708,
      "step": 6650
    },
    {
      "epoch": 2.96,
      "learning_rate": 6.666666666666667e-07,
      "loss": 0.1736,
      "step": 6660
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.925925925925926e-07,
      "loss": 0.1654,
      "step": 6670
    },
    {
      "epoch": 2.97,
      "eval_loss": 0.24077662825584412,
      "eval_runtime": 7.2025,
      "eval_samples_per_second": 138.98,
      "eval_steps_per_second": 34.849,
      "step": 6675
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.185185185185186e-07,
      "loss": 0.1993,
      "step": 6680
    },
    {
      "epoch": 2.97,
      "learning_rate": 4.444444444444445e-07,
      "loss": 0.1852,
      "step": 6690
    },
    {
      "epoch": 2.98,
      "learning_rate": 3.703703703703704e-07,
      "loss": 0.1955,
      "step": 6700
    },
    {
      "epoch": 2.98,
      "eval_loss": 0.24076078832149506,
      "eval_runtime": 7.1929,
      "eval_samples_per_second": 139.165,
      "eval_steps_per_second": 34.895,
      "step": 6700
    },
    {
      "epoch": 2.98,
      "learning_rate": 2.962962962962963e-07,
      "loss": 0.1923,
      "step": 6710
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 0.2142,
      "step": 6720
    },
    {
      "epoch": 2.99,
      "eval_loss": 0.24076122045516968,
      "eval_runtime": 7.1938,
      "eval_samples_per_second": 139.147,
      "eval_steps_per_second": 34.891,
      "step": 6725
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.4814814814814815e-07,
      "loss": 0.2464,
      "step": 6730
    },
    {
      "epoch": 3.0,
      "learning_rate": 7.407407407407407e-08,
      "loss": 0.2322,
      "step": 6740
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.0,
      "loss": 0.1781,
      "step": 6750
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.24075740575790405,
      "eval_runtime": 7.2037,
      "eval_samples_per_second": 138.956,
      "eval_steps_per_second": 34.843,
      "step": 6750
    }
  ],
  "max_steps": 6750,
  "num_train_epochs": 3,
  "total_flos": 915255853056000.0,
  "trial_name": null,
  "trial_params": null
}