{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.433962264150944,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 1e-05,
      "loss": 0.9816,
      "step": 10
    },
    {
      "epoch": 0.05,
      "learning_rate": 2e-05,
      "loss": 0.9819,
      "step": 20
    },
    {
      "epoch": 0.07,
      "learning_rate": 3e-05,
      "loss": 0.7618,
      "step": 30
    },
    {
      "epoch": 0.09,
      "learning_rate": 4e-05,
      "loss": 0.71,
      "step": 40
    },
    {
      "epoch": 0.12,
      "learning_rate": 5e-05,
      "loss": 0.5227,
      "step": 50
    },
    {
      "epoch": 0.14,
      "learning_rate": 6e-05,
      "loss": 0.4715,
      "step": 60
    },
    {
      "epoch": 0.17,
      "learning_rate": 7e-05,
      "loss": 0.4725,
      "step": 70
    },
    {
      "epoch": 0.19,
      "learning_rate": 8e-05,
      "loss": 0.4478,
      "step": 80
    },
    {
      "epoch": 0.21,
      "learning_rate": 9e-05,
      "loss": 0.4305,
      "step": 90
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.0001,
      "loss": 0.4244,
      "step": 100
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.97584541062802e-05,
      "loss": 0.4307,
      "step": 110
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.951690821256039e-05,
      "loss": 0.3962,
      "step": 120
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.927536231884058e-05,
      "loss": 0.3897,
      "step": 130
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.903381642512077e-05,
      "loss": 0.392,
      "step": 140
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.879227053140096e-05,
      "loss": 0.3898,
      "step": 150
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.855072463768117e-05,
      "loss": 0.3678,
      "step": 160
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.830917874396136e-05,
      "loss": 0.3695,
      "step": 170
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.806763285024155e-05,
      "loss": 0.3813,
      "step": 180
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.782608695652174e-05,
      "loss": 0.3568,
      "step": 190
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.758454106280194e-05,
      "loss": 0.3622,
      "step": 200
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.734299516908213e-05,
      "loss": 0.3795,
      "step": 210
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.710144927536232e-05,
      "loss": 0.3565,
      "step": 220
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.685990338164251e-05,
      "loss": 0.347,
      "step": 230
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.66183574879227e-05,
      "loss": 0.3329,
      "step": 240
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.63768115942029e-05,
      "loss": 0.3425,
      "step": 250
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.61352657004831e-05,
      "loss": 0.3395,
      "step": 260
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.589371980676329e-05,
      "loss": 0.364,
      "step": 270
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.565217391304348e-05,
      "loss": 0.3454,
      "step": 280
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.541062801932368e-05,
      "loss": 0.3261,
      "step": 290
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.516908212560387e-05,
      "loss": 0.3426,
      "step": 300
    },
    {
      "epoch": 0.73,
      "learning_rate": 9.492753623188406e-05,
      "loss": 0.3217,
      "step": 310
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.468599033816425e-05,
      "loss": 0.3348,
      "step": 320
    },
    {
      "epoch": 0.78,
      "learning_rate": 9.444444444444444e-05,
      "loss": 0.3169,
      "step": 330
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.420289855072463e-05,
      "loss": 0.3537,
      "step": 340
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.396135265700483e-05,
      "loss": 0.2869,
      "step": 350
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.371980676328503e-05,
      "loss": 0.3426,
      "step": 360
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.347826086956522e-05,
      "loss": 0.3266,
      "step": 370
    },
    {
      "epoch": 0.9,
      "learning_rate": 9.323671497584541e-05,
      "loss": 0.3355,
      "step": 380
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.29951690821256e-05,
      "loss": 0.3083,
      "step": 390
    },
    {
      "epoch": 0.94,
      "learning_rate": 9.27536231884058e-05,
      "loss": 0.3047,
      "step": 400
    },
    {
      "epoch": 0.97,
      "learning_rate": 9.251207729468599e-05,
      "loss": 0.3387,
      "step": 410
    },
    {
      "epoch": 0.99,
      "learning_rate": 9.227053140096618e-05,
      "loss": 0.3297,
      "step": 420
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.3164224922657013,
      "eval_runtime": 9.557,
      "eval_samples_per_second": 78.79,
      "eval_steps_per_second": 5.022,
      "step": 424
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.202898550724637e-05,
      "loss": 0.3035,
      "step": 430
    },
    {
      "epoch": 1.04,
      "learning_rate": 9.178743961352657e-05,
      "loss": 0.2824,
      "step": 440
    },
    {
      "epoch": 1.06,
      "learning_rate": 9.154589371980676e-05,
      "loss": 0.3045,
      "step": 450
    },
    {
      "epoch": 1.08,
      "learning_rate": 9.130434782608696e-05,
      "loss": 0.2843,
      "step": 460
    },
    {
      "epoch": 1.11,
      "learning_rate": 9.106280193236715e-05,
      "loss": 0.2834,
      "step": 470
    },
    {
      "epoch": 1.13,
      "learning_rate": 9.082125603864735e-05,
      "loss": 0.2681,
      "step": 480
    },
    {
      "epoch": 1.16,
      "learning_rate": 9.057971014492754e-05,
      "loss": 0.266,
      "step": 490
    },
    {
      "epoch": 1.18,
      "learning_rate": 9.033816425120773e-05,
      "loss": 0.2939,
      "step": 500
    },
    {
      "epoch": 1.2,
      "learning_rate": 9.009661835748792e-05,
      "loss": 0.276,
      "step": 510
    },
    {
      "epoch": 1.23,
      "learning_rate": 8.985507246376813e-05,
      "loss": 0.2607,
      "step": 520
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.961352657004832e-05,
      "loss": 0.2761,
      "step": 530
    },
    {
      "epoch": 1.27,
      "learning_rate": 8.937198067632851e-05,
      "loss": 0.2821,
      "step": 540
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.91304347826087e-05,
      "loss": 0.2651,
      "step": 550
    },
    {
      "epoch": 1.32,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.3199,
      "step": 560
    },
    {
      "epoch": 1.34,
      "learning_rate": 8.864734299516908e-05,
      "loss": 0.2759,
      "step": 570
    },
    {
      "epoch": 1.37,
      "learning_rate": 8.840579710144929e-05,
      "loss": 0.2689,
      "step": 580
    },
    {
      "epoch": 1.39,
      "learning_rate": 8.816425120772948e-05,
      "loss": 0.2988,
      "step": 590
    },
    {
      "epoch": 1.42,
      "learning_rate": 8.792270531400967e-05,
      "loss": 0.2558,
      "step": 600
    },
    {
      "epoch": 1.44,
      "learning_rate": 8.768115942028987e-05,
      "loss": 0.279,
      "step": 610
    },
    {
      "epoch": 1.46,
      "learning_rate": 8.743961352657006e-05,
      "loss": 0.2852,
      "step": 620
    },
    {
      "epoch": 1.49,
      "learning_rate": 8.719806763285025e-05,
      "loss": 0.2839,
      "step": 630
    },
    {
      "epoch": 1.51,
      "learning_rate": 8.695652173913044e-05,
      "loss": 0.269,
      "step": 640
    },
    {
      "epoch": 1.53,
      "learning_rate": 8.671497584541063e-05,
      "loss": 0.2648,
      "step": 650
    },
    {
      "epoch": 1.56,
      "learning_rate": 8.647342995169082e-05,
      "loss": 0.2712,
      "step": 660
    },
    {
      "epoch": 1.58,
      "learning_rate": 8.623188405797103e-05,
      "loss": 0.2652,
      "step": 670
    },
    {
      "epoch": 1.6,
      "learning_rate": 8.599033816425122e-05,
      "loss": 0.2512,
      "step": 680
    },
    {
      "epoch": 1.63,
      "learning_rate": 8.574879227053141e-05,
      "loss": 0.2691,
      "step": 690
    },
    {
      "epoch": 1.65,
      "learning_rate": 8.55072463768116e-05,
      "loss": 0.2504,
      "step": 700
    },
    {
      "epoch": 1.67,
      "learning_rate": 8.52657004830918e-05,
      "loss": 0.2584,
      "step": 710
    },
    {
      "epoch": 1.7,
      "learning_rate": 8.502415458937199e-05,
      "loss": 0.2536,
      "step": 720
    },
    {
      "epoch": 1.72,
      "learning_rate": 8.478260869565218e-05,
      "loss": 0.2604,
      "step": 730
    },
    {
      "epoch": 1.75,
      "learning_rate": 8.454106280193237e-05,
      "loss": 0.2818,
      "step": 740
    },
    {
      "epoch": 1.77,
      "learning_rate": 8.429951690821256e-05,
      "loss": 0.2766,
      "step": 750
    },
    {
      "epoch": 1.79,
      "learning_rate": 8.405797101449276e-05,
      "loss": 0.2697,
      "step": 760
    },
    {
      "epoch": 1.82,
      "learning_rate": 8.381642512077295e-05,
      "loss": 0.2457,
      "step": 770
    },
    {
      "epoch": 1.84,
      "learning_rate": 8.357487922705315e-05,
      "loss": 0.2665,
      "step": 780
    },
    {
      "epoch": 1.86,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.3055,
      "step": 790
    },
    {
      "epoch": 1.89,
      "learning_rate": 8.309178743961354e-05,
      "loss": 0.2733,
      "step": 800
    },
    {
      "epoch": 1.91,
      "learning_rate": 8.285024154589373e-05,
      "loss": 0.2604,
      "step": 810
    },
    {
      "epoch": 1.93,
      "learning_rate": 8.260869565217392e-05,
      "loss": 0.2689,
      "step": 820
    },
    {
      "epoch": 1.96,
      "learning_rate": 8.236714975845411e-05,
      "loss": 0.2735,
      "step": 830
    },
    {
      "epoch": 1.98,
      "learning_rate": 8.21256038647343e-05,
      "loss": 0.2795,
      "step": 840
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.30382633209228516,
      "eval_runtime": 9.4692,
      "eval_samples_per_second": 79.521,
      "eval_steps_per_second": 5.069,
      "step": 848
    },
    {
      "epoch": 2.0,
      "learning_rate": 8.18840579710145e-05,
      "loss": 0.2476,
      "step": 850
    },
    {
      "epoch": 2.03,
      "learning_rate": 8.164251207729469e-05,
      "loss": 0.2192,
      "step": 860
    },
    {
      "epoch": 2.05,
      "learning_rate": 8.140096618357488e-05,
      "loss": 0.2083,
      "step": 870
    },
    {
      "epoch": 2.08,
      "learning_rate": 8.115942028985508e-05,
      "loss": 0.2206,
      "step": 880
    },
    {
      "epoch": 2.1,
      "learning_rate": 8.091787439613528e-05,
      "loss": 0.2209,
      "step": 890
    },
    {
      "epoch": 2.12,
      "learning_rate": 8.067632850241547e-05,
      "loss": 0.2158,
      "step": 900
    },
    {
      "epoch": 2.15,
      "learning_rate": 8.043478260869566e-05,
      "loss": 0.2105,
      "step": 910
    },
    {
      "epoch": 2.17,
      "learning_rate": 8.019323671497585e-05,
      "loss": 0.2106,
      "step": 920
    },
    {
      "epoch": 2.19,
      "learning_rate": 7.995169082125604e-05,
      "loss": 0.2292,
      "step": 930
    },
    {
      "epoch": 2.22,
      "learning_rate": 7.971014492753623e-05,
      "loss": 0.2096,
      "step": 940
    },
    {
      "epoch": 2.24,
      "learning_rate": 7.946859903381643e-05,
      "loss": 0.2224,
      "step": 950
    },
    {
      "epoch": 2.26,
      "learning_rate": 7.922705314009662e-05,
      "loss": 0.2241,
      "step": 960
    },
    {
      "epoch": 2.29,
      "learning_rate": 7.898550724637681e-05,
      "loss": 0.2315,
      "step": 970
    },
    {
      "epoch": 2.31,
      "learning_rate": 7.874396135265701e-05,
      "loss": 0.2128,
      "step": 980
    },
    {
      "epoch": 2.33,
      "learning_rate": 7.85024154589372e-05,
      "loss": 0.2232,
      "step": 990
    },
    {
      "epoch": 2.36,
      "learning_rate": 7.82608695652174e-05,
      "loss": 0.212,
      "step": 1000
    },
    {
      "epoch": 2.38,
      "learning_rate": 7.801932367149759e-05,
      "loss": 0.2188,
      "step": 1010
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.2131,
      "step": 1020
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.753623188405797e-05,
      "loss": 0.2223,
      "step": 1030
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.729468599033817e-05,
      "loss": 0.2134,
      "step": 1040
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.705314009661836e-05,
      "loss": 0.213,
      "step": 1050
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.681159420289855e-05,
      "loss": 0.2102,
      "step": 1060
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.657004830917874e-05,
      "loss": 0.228,
      "step": 1070
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.632850241545893e-05,
      "loss": 0.2274,
      "step": 1080
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.608695652173914e-05,
      "loss": 0.2171,
      "step": 1090
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.584541062801933e-05,
      "loss": 0.2035,
      "step": 1100
    },
    {
      "epoch": 2.62,
      "learning_rate": 7.560386473429952e-05,
      "loss": 0.2108,
      "step": 1110
    },
    {
      "epoch": 2.64,
      "learning_rate": 7.536231884057971e-05,
      "loss": 0.2259,
      "step": 1120
    },
    {
      "epoch": 2.67,
      "learning_rate": 7.51207729468599e-05,
      "loss": 0.2158,
      "step": 1130
    },
    {
      "epoch": 2.69,
      "learning_rate": 7.48792270531401e-05,
      "loss": 0.2257,
      "step": 1140
    },
    {
      "epoch": 2.71,
      "learning_rate": 7.463768115942029e-05,
      "loss": 0.2086,
      "step": 1150
    },
    {
      "epoch": 2.74,
      "learning_rate": 7.439613526570048e-05,
      "loss": 0.199,
      "step": 1160
    },
    {
      "epoch": 2.76,
      "learning_rate": 7.415458937198067e-05,
      "loss": 0.2302,
      "step": 1170
    },
    {
      "epoch": 2.78,
      "learning_rate": 7.391304347826086e-05,
      "loss": 0.2168,
      "step": 1180
    },
    {
      "epoch": 2.81,
      "learning_rate": 7.367149758454107e-05,
      "loss": 0.2138,
      "step": 1190
    },
    {
      "epoch": 2.83,
      "learning_rate": 7.342995169082126e-05,
      "loss": 0.2133,
      "step": 1200
    },
    {
      "epoch": 2.85,
      "learning_rate": 7.318840579710145e-05,
      "loss": 0.2129,
      "step": 1210
    },
    {
      "epoch": 2.88,
      "learning_rate": 7.294685990338164e-05,
      "loss": 0.2321,
      "step": 1220
    },
    {
      "epoch": 2.9,
      "learning_rate": 7.270531400966184e-05,
      "loss": 0.2162,
      "step": 1230
    },
    {
      "epoch": 2.92,
      "learning_rate": 7.246376811594203e-05,
      "loss": 0.2105,
      "step": 1240
    },
    {
      "epoch": 2.95,
      "learning_rate": 7.222222222222222e-05,
      "loss": 0.2335,
      "step": 1250
    },
    {
      "epoch": 2.97,
      "learning_rate": 7.198067632850241e-05,
      "loss": 0.1969,
      "step": 1260
    },
    {
      "epoch": 3.0,
      "learning_rate": 7.17391304347826e-05,
      "loss": 0.2199,
      "step": 1270
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.3284262716770172,
      "eval_runtime": 10.1976,
      "eval_samples_per_second": 73.841,
      "eval_steps_per_second": 4.707,
      "step": 1272
    },
    {
      "epoch": 3.02,
      "learning_rate": 7.14975845410628e-05,
      "loss": 0.1854,
      "step": 1280
    },
    {
      "epoch": 3.04,
      "learning_rate": 7.1256038647343e-05,
      "loss": 0.1687,
      "step": 1290
    },
    {
      "epoch": 3.07,
      "learning_rate": 7.101449275362319e-05,
      "loss": 0.1867,
      "step": 1300
    },
    {
      "epoch": 3.09,
      "learning_rate": 7.077294685990338e-05,
      "loss": 0.1816,
      "step": 1310
    },
    {
      "epoch": 3.11,
      "learning_rate": 7.053140096618357e-05,
      "loss": 0.169,
      "step": 1320
    },
    {
      "epoch": 3.14,
      "learning_rate": 7.028985507246377e-05,
      "loss": 0.1747,
      "step": 1330
    },
    {
      "epoch": 3.16,
      "learning_rate": 7.004830917874396e-05,
      "loss": 0.1823,
      "step": 1340
    },
    {
      "epoch": 3.18,
      "learning_rate": 6.980676328502416e-05,
      "loss": 0.1873,
      "step": 1350
    },
    {
      "epoch": 3.21,
      "learning_rate": 6.956521739130436e-05,
      "loss": 0.1716,
      "step": 1360
    },
    {
      "epoch": 3.23,
      "learning_rate": 6.932367149758455e-05,
      "loss": 0.186,
      "step": 1370
    },
    {
      "epoch": 3.25,
      "learning_rate": 6.908212560386474e-05,
      "loss": 0.1822,
      "step": 1380
    },
    {
      "epoch": 3.28,
      "learning_rate": 6.884057971014493e-05,
      "loss": 0.1741,
      "step": 1390
    },
    {
      "epoch": 3.3,
      "learning_rate": 6.859903381642512e-05,
      "loss": 0.1782,
      "step": 1400
    },
    {
      "epoch": 3.33,
      "learning_rate": 6.835748792270533e-05,
      "loss": 0.1777,
      "step": 1410
    },
    {
      "epoch": 3.35,
      "learning_rate": 6.811594202898552e-05,
      "loss": 0.1921,
      "step": 1420
    },
    {
      "epoch": 3.37,
      "learning_rate": 6.787439613526571e-05,
      "loss": 0.1821,
      "step": 1430
    },
    {
      "epoch": 3.4,
      "learning_rate": 6.76328502415459e-05,
      "loss": 0.162,
      "step": 1440
    },
    {
      "epoch": 3.42,
      "learning_rate": 6.73913043478261e-05,
      "loss": 0.1722,
      "step": 1450
    },
    {
      "epoch": 3.44,
      "learning_rate": 6.714975845410629e-05,
      "loss": 0.1702,
      "step": 1460
    },
    {
      "epoch": 3.47,
      "learning_rate": 6.690821256038648e-05,
      "loss": 0.1685,
      "step": 1470
    },
    {
      "epoch": 3.49,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2024,
      "step": 1480
    },
    {
      "epoch": 3.51,
      "learning_rate": 6.642512077294686e-05,
      "loss": 0.1785,
      "step": 1490
    },
    {
      "epoch": 3.54,
      "learning_rate": 6.618357487922707e-05,
      "loss": 0.1703,
      "step": 1500
    },
    {
      "epoch": 3.56,
      "learning_rate": 6.594202898550726e-05,
      "loss": 0.1879,
      "step": 1510
    },
    {
      "epoch": 3.58,
      "learning_rate": 6.570048309178745e-05,
      "loss": 0.1837,
      "step": 1520
    },
    {
      "epoch": 3.61,
      "learning_rate": 6.545893719806764e-05,
      "loss": 0.1656,
      "step": 1530
    },
    {
      "epoch": 3.63,
      "learning_rate": 6.521739130434783e-05,
      "loss": 0.1669,
      "step": 1540
    },
    {
      "epoch": 3.66,
      "learning_rate": 6.497584541062803e-05,
      "loss": 0.173,
      "step": 1550
    },
    {
      "epoch": 3.68,
      "learning_rate": 6.473429951690822e-05,
      "loss": 0.1958,
      "step": 1560
    },
    {
      "epoch": 3.7,
      "learning_rate": 6.449275362318841e-05,
      "loss": 0.1908,
      "step": 1570
    },
    {
      "epoch": 3.73,
      "learning_rate": 6.42512077294686e-05,
      "loss": 0.1969,
      "step": 1580
    },
    {
      "epoch": 3.75,
      "learning_rate": 6.400966183574879e-05,
      "loss": 0.1663,
      "step": 1590
    },
    {
      "epoch": 3.77,
      "learning_rate": 6.376811594202898e-05,
      "loss": 0.1939,
      "step": 1600
    },
    {
      "epoch": 3.8,
      "learning_rate": 6.352657004830919e-05,
      "loss": 0.1861,
      "step": 1610
    },
    {
      "epoch": 3.82,
      "learning_rate": 6.328502415458938e-05,
      "loss": 0.1865,
      "step": 1620
    },
    {
      "epoch": 3.84,
      "learning_rate": 6.304347826086957e-05,
      "loss": 0.1849,
      "step": 1630
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.280193236714976e-05,
      "loss": 0.1796,
      "step": 1640
    },
    {
      "epoch": 3.89,
      "learning_rate": 6.256038647342996e-05,
      "loss": 0.1725,
      "step": 1650
    },
    {
      "epoch": 3.92,
      "learning_rate": 6.231884057971015e-05,
      "loss": 0.1884,
      "step": 1660
    },
    {
      "epoch": 3.94,
      "learning_rate": 6.207729468599034e-05,
      "loss": 0.166,
      "step": 1670
    },
    {
      "epoch": 3.96,
      "learning_rate": 6.183574879227053e-05,
      "loss": 0.1915,
      "step": 1680
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.159420289855072e-05,
      "loss": 0.1767,
      "step": 1690
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.385652631521225,
      "eval_runtime": 9.485,
      "eval_samples_per_second": 79.388,
      "eval_steps_per_second": 5.061,
      "step": 1696
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.135265700483092e-05,
      "loss": 0.1665,
      "step": 1700
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.111111111111112e-05,
      "loss": 0.1519,
      "step": 1710
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.086956521739131e-05,
      "loss": 0.1573,
      "step": 1720
    },
    {
      "epoch": 4.08,
      "learning_rate": 6.0628019323671504e-05,
      "loss": 0.1593,
      "step": 1730
    },
    {
      "epoch": 4.1,
      "learning_rate": 6.0386473429951696e-05,
      "loss": 0.151,
      "step": 1740
    },
    {
      "epoch": 4.13,
      "learning_rate": 6.014492753623189e-05,
      "loss": 0.1566,
      "step": 1750
    },
    {
      "epoch": 4.15,
      "learning_rate": 5.990338164251208e-05,
      "loss": 0.1563,
      "step": 1760
    },
    {
      "epoch": 4.17,
      "learning_rate": 5.966183574879227e-05,
      "loss": 0.1454,
      "step": 1770
    },
    {
      "epoch": 4.2,
      "learning_rate": 5.942028985507246e-05,
      "loss": 0.1555,
      "step": 1780
    },
    {
      "epoch": 4.22,
      "learning_rate": 5.9178743961352655e-05,
      "loss": 0.158,
      "step": 1790
    },
    {
      "epoch": 4.25,
      "learning_rate": 5.8937198067632847e-05,
      "loss": 0.1574,
      "step": 1800
    },
    {
      "epoch": 4.27,
      "learning_rate": 5.869565217391305e-05,
      "loss": 0.1522,
      "step": 1810
    },
    {
      "epoch": 4.29,
      "learning_rate": 5.8454106280193244e-05,
      "loss": 0.1467,
      "step": 1820
    },
    {
      "epoch": 4.32,
      "learning_rate": 5.8212560386473435e-05,
      "loss": 0.1475,
      "step": 1830
    },
    {
      "epoch": 4.34,
      "learning_rate": 5.797101449275363e-05,
      "loss": 0.1441,
      "step": 1840
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.772946859903382e-05,
      "loss": 0.1558,
      "step": 1850
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.748792270531401e-05,
      "loss": 0.1542,
      "step": 1860
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.72463768115942e-05,
      "loss": 0.1565,
      "step": 1870
    },
    {
      "epoch": 4.43,
      "learning_rate": 5.7004830917874394e-05,
      "loss": 0.1586,
      "step": 1880
    },
    {
      "epoch": 4.46,
      "learning_rate": 5.6763285024154586e-05,
      "loss": 0.1526,
      "step": 1890
    },
    {
      "epoch": 4.48,
      "learning_rate": 5.652173913043478e-05,
      "loss": 0.15,
      "step": 1900
    },
    {
      "epoch": 4.5,
      "learning_rate": 5.6280193236714976e-05,
      "loss": 0.1631,
      "step": 1910
    },
    {
      "epoch": 4.53,
      "learning_rate": 5.6038647342995175e-05,
      "loss": 0.1516,
      "step": 1920
    },
    {
      "epoch": 4.55,
      "learning_rate": 5.5797101449275366e-05,
      "loss": 0.1539,
      "step": 1930
    },
    {
      "epoch": 4.58,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.1689,
      "step": 1940
    },
    {
      "epoch": 4.6,
      "learning_rate": 5.531400966183575e-05,
      "loss": 0.1572,
      "step": 1950
    },
    {
      "epoch": 4.62,
      "learning_rate": 5.507246376811594e-05,
      "loss": 0.1523,
      "step": 1960
    },
    {
      "epoch": 4.65,
      "learning_rate": 5.483091787439614e-05,
      "loss": 0.1626,
      "step": 1970
    },
    {
      "epoch": 4.67,
      "learning_rate": 5.458937198067633e-05,
      "loss": 0.1597,
      "step": 1980
    },
    {
      "epoch": 4.69,
      "learning_rate": 5.4347826086956524e-05,
      "loss": 0.1515,
      "step": 1990
    },
    {
      "epoch": 4.72,
      "learning_rate": 5.4106280193236716e-05,
      "loss": 0.1573,
      "step": 2000
    },
    {
      "epoch": 4.74,
      "learning_rate": 5.386473429951691e-05,
      "loss": 0.1554,
      "step": 2010
    },
    {
      "epoch": 4.76,
      "learning_rate": 5.3623188405797106e-05,
      "loss": 0.161,
      "step": 2020
    },
    {
      "epoch": 4.79,
      "learning_rate": 5.3381642512077304e-05,
      "loss": 0.1503,
      "step": 2030
    },
    {
      "epoch": 4.81,
      "learning_rate": 5.3140096618357496e-05,
      "loss": 0.1548,
      "step": 2040
    },
    {
      "epoch": 4.83,
      "learning_rate": 5.289855072463769e-05,
      "loss": 0.1529,
      "step": 2050
    },
    {
      "epoch": 4.86,
      "learning_rate": 5.265700483091788e-05,
      "loss": 0.1524,
      "step": 2060
    },
    {
      "epoch": 4.88,
      "learning_rate": 5.241545893719807e-05,
      "loss": 0.1674,
      "step": 2070
    },
    {
      "epoch": 4.91,
      "learning_rate": 5.217391304347826e-05,
      "loss": 0.1633,
      "step": 2080
    },
    {
      "epoch": 4.93,
      "learning_rate": 5.1932367149758455e-05,
      "loss": 0.1593,
      "step": 2090
    },
    {
      "epoch": 4.95,
      "learning_rate": 5.1690821256038647e-05,
      "loss": 0.1502,
      "step": 2100
    },
    {
      "epoch": 4.98,
      "learning_rate": 5.144927536231884e-05,
      "loss": 0.1546,
      "step": 2110
    },
    {
      "epoch": 5.0,
      "learning_rate": 5.1207729468599044e-05,
      "loss": 0.1583,
      "step": 2120
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.4561803936958313,
      "eval_runtime": 9.5555,
      "eval_samples_per_second": 78.803,
      "eval_steps_per_second": 5.023,
      "step": 2120
    },
    {
      "epoch": 5.02,
      "learning_rate": 5.0966183574879235e-05,
      "loss": 0.1428,
      "step": 2130
    },
    {
      "epoch": 5.05,
      "learning_rate": 5.072463768115943e-05,
      "loss": 0.1389,
      "step": 2140
    },
    {
      "epoch": 5.07,
      "learning_rate": 5.048309178743962e-05,
      "loss": 0.1449,
      "step": 2150
    },
    {
      "epoch": 5.09,
      "learning_rate": 5.024154589371981e-05,
      "loss": 0.1427,
      "step": 2160
    },
    {
      "epoch": 5.12,
      "learning_rate": 5e-05,
      "loss": 0.1359,
      "step": 2170
    },
    {
      "epoch": 5.14,
      "learning_rate": 4.9758454106280194e-05,
      "loss": 0.158,
      "step": 2180
    },
    {
      "epoch": 5.17,
      "learning_rate": 4.9516908212560386e-05,
      "loss": 0.1368,
      "step": 2190
    },
    {
      "epoch": 5.19,
      "learning_rate": 4.9275362318840584e-05,
      "loss": 0.1316,
      "step": 2200
    },
    {
      "epoch": 5.21,
      "learning_rate": 4.9033816425120776e-05,
      "loss": 0.1389,
      "step": 2210
    },
    {
      "epoch": 5.24,
      "learning_rate": 4.879227053140097e-05,
      "loss": 0.138,
      "step": 2220
    },
    {
      "epoch": 5.26,
      "learning_rate": 4.855072463768116e-05,
      "loss": 0.1407,
      "step": 2230
    },
    {
      "epoch": 5.28,
      "learning_rate": 4.830917874396135e-05,
      "loss": 0.1475,
      "step": 2240
    },
    {
      "epoch": 5.31,
      "learning_rate": 4.806763285024155e-05,
      "loss": 0.148,
      "step": 2250
    },
    {
      "epoch": 5.33,
      "learning_rate": 4.782608695652174e-05,
      "loss": 0.138,
      "step": 2260
    },
    {
      "epoch": 5.35,
      "learning_rate": 4.7584541062801933e-05,
      "loss": 0.1375,
      "step": 2270
    },
    {
      "epoch": 5.38,
      "learning_rate": 4.7342995169082125e-05,
      "loss": 0.1407,
      "step": 2280
    },
    {
      "epoch": 5.4,
      "learning_rate": 4.710144927536232e-05,
      "loss": 0.1351,
      "step": 2290
    },
    {
      "epoch": 5.42,
      "learning_rate": 4.6859903381642516e-05,
      "loss": 0.1409,
      "step": 2300
    },
    {
      "epoch": 5.45,
      "learning_rate": 4.661835748792271e-05,
      "loss": 0.1366,
      "step": 2310
    },
    {
      "epoch": 5.47,
      "learning_rate": 4.63768115942029e-05,
      "loss": 0.1389,
      "step": 2320
    },
    {
      "epoch": 5.5,
      "learning_rate": 4.613526570048309e-05,
      "loss": 0.1435,
      "step": 2330
    },
    {
      "epoch": 5.52,
      "learning_rate": 4.589371980676328e-05,
      "loss": 0.141,
      "step": 2340
    },
    {
      "epoch": 5.54,
      "learning_rate": 4.565217391304348e-05,
      "loss": 0.1468,
      "step": 2350
    },
    {
      "epoch": 5.57,
      "learning_rate": 4.541062801932367e-05,
      "loss": 0.1417,
      "step": 2360
    },
    {
      "epoch": 5.59,
      "learning_rate": 4.5169082125603865e-05,
      "loss": 0.1359,
      "step": 2370
    },
    {
      "epoch": 5.61,
      "learning_rate": 4.492753623188406e-05,
      "loss": 0.1362,
      "step": 2380
    },
    {
      "epoch": 5.64,
      "learning_rate": 4.4685990338164255e-05,
      "loss": 0.1494,
      "step": 2390
    },
    {
      "epoch": 5.66,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.1431,
      "step": 2400
    },
    {
      "epoch": 5.68,
      "learning_rate": 4.4202898550724645e-05,
      "loss": 0.1364,
      "step": 2410
    },
    {
      "epoch": 5.71,
      "learning_rate": 4.396135265700484e-05,
      "loss": 0.1488,
      "step": 2420
    },
    {
      "epoch": 5.73,
      "learning_rate": 4.371980676328503e-05,
      "loss": 0.1446,
      "step": 2430
    },
    {
      "epoch": 5.75,
      "learning_rate": 4.347826086956522e-05,
      "loss": 0.1368,
      "step": 2440
    },
    {
      "epoch": 5.78,
      "learning_rate": 4.323671497584541e-05,
      "loss": 0.1374,
      "step": 2450
    },
    {
      "epoch": 5.8,
      "learning_rate": 4.299516908212561e-05,
      "loss": 0.1417,
      "step": 2460
    },
    {
      "epoch": 5.83,
      "learning_rate": 4.27536231884058e-05,
      "loss": 0.1379,
      "step": 2470
    },
    {
      "epoch": 5.85,
      "learning_rate": 4.2512077294685994e-05,
      "loss": 0.138,
      "step": 2480
    },
    {
      "epoch": 5.87,
      "learning_rate": 4.2270531400966186e-05,
      "loss": 0.1335,
      "step": 2490
    },
    {
      "epoch": 5.9,
      "learning_rate": 4.202898550724638e-05,
      "loss": 0.1395,
      "step": 2500
    },
    {
      "epoch": 5.92,
      "learning_rate": 4.1787439613526576e-05,
      "loss": 0.1364,
      "step": 2510
    },
    {
      "epoch": 5.94,
      "learning_rate": 4.154589371980677e-05,
      "loss": 0.1511,
      "step": 2520
    },
    {
      "epoch": 5.97,
      "learning_rate": 4.130434782608696e-05,
      "loss": 0.1369,
      "step": 2530
    },
    {
      "epoch": 5.99,
      "learning_rate": 4.106280193236715e-05,
      "loss": 0.139,
      "step": 2540
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.49535447359085083,
      "eval_runtime": 9.4913,
      "eval_samples_per_second": 79.336,
      "eval_steps_per_second": 5.057,
      "step": 2544
    },
    {
      "epoch": 6.01,
      "learning_rate": 4.082125603864734e-05,
      "loss": 0.1277,
      "step": 2550
    },
    {
      "epoch": 6.04,
      "learning_rate": 4.057971014492754e-05,
      "loss": 0.1344,
      "step": 2560
    },
    {
      "epoch": 6.06,
      "learning_rate": 4.0338164251207733e-05,
      "loss": 0.1268,
      "step": 2570
    },
    {
      "epoch": 6.08,
      "learning_rate": 4.0096618357487925e-05,
      "loss": 0.1279,
      "step": 2580
    },
    {
      "epoch": 6.11,
      "learning_rate": 3.985507246376812e-05,
      "loss": 0.129,
      "step": 2590
    },
    {
      "epoch": 6.13,
      "learning_rate": 3.961352657004831e-05,
      "loss": 0.1313,
      "step": 2600
    },
    {
      "epoch": 6.16,
      "learning_rate": 3.937198067632851e-05,
      "loss": 0.1259,
      "step": 2610
    },
    {
      "epoch": 6.18,
      "learning_rate": 3.91304347826087e-05,
      "loss": 0.1347,
      "step": 2620
    },
    {
      "epoch": 6.2,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.1224,
      "step": 2630
    },
    {
      "epoch": 6.23,
      "learning_rate": 3.864734299516908e-05,
      "loss": 0.1256,
      "step": 2640
    },
    {
      "epoch": 6.25,
      "learning_rate": 3.8405797101449274e-05,
      "loss": 0.1303,
      "step": 2650
    },
    {
      "epoch": 6.27,
      "learning_rate": 3.8164251207729466e-05,
      "loss": 0.1326,
      "step": 2660
    },
    {
      "epoch": 6.3,
      "learning_rate": 3.7922705314009665e-05,
      "loss": 0.1325,
      "step": 2670
    },
    {
      "epoch": 6.32,
      "learning_rate": 3.7681159420289856e-05,
      "loss": 0.1339,
      "step": 2680
    },
    {
      "epoch": 6.34,
      "learning_rate": 3.743961352657005e-05,
      "loss": 0.127,
      "step": 2690
    },
    {
      "epoch": 6.37,
      "learning_rate": 3.719806763285024e-05,
      "loss": 0.1352,
      "step": 2700
    },
    {
      "epoch": 6.39,
      "learning_rate": 3.695652173913043e-05,
      "loss": 0.1261,
      "step": 2710
    },
    {
      "epoch": 6.42,
      "learning_rate": 3.671497584541063e-05,
      "loss": 0.1277,
      "step": 2720
    },
    {
      "epoch": 6.44,
      "learning_rate": 3.647342995169082e-05,
      "loss": 0.1296,
      "step": 2730
    },
    {
      "epoch": 6.46,
      "learning_rate": 3.6231884057971014e-05,
      "loss": 0.127,
      "step": 2740
    },
    {
      "epoch": 6.49,
      "learning_rate": 3.5990338164251205e-05,
      "loss": 0.1327,
      "step": 2750
    },
    {
      "epoch": 6.51,
      "learning_rate": 3.57487922705314e-05,
      "loss": 0.1286,
      "step": 2760
    },
    {
      "epoch": 6.53,
      "learning_rate": 3.5507246376811596e-05,
      "loss": 0.1346,
      "step": 2770
    },
    {
      "epoch": 6.56,
      "learning_rate": 3.526570048309179e-05,
      "loss": 0.1276,
      "step": 2780
    },
    {
      "epoch": 6.58,
      "learning_rate": 3.502415458937198e-05,
      "loss": 0.124,
      "step": 2790
    },
    {
      "epoch": 6.6,
      "learning_rate": 3.478260869565218e-05,
      "loss": 0.1338,
      "step": 2800
    },
    {
      "epoch": 6.63,
      "learning_rate": 3.454106280193237e-05,
      "loss": 0.1276,
      "step": 2810
    },
    {
      "epoch": 6.65,
      "learning_rate": 3.429951690821256e-05,
      "loss": 0.127,
      "step": 2820
    },
    {
      "epoch": 6.67,
      "learning_rate": 3.405797101449276e-05,
      "loss": 0.1262,
      "step": 2830
    },
    {
      "epoch": 6.7,
      "learning_rate": 3.381642512077295e-05,
      "loss": 0.1239,
      "step": 2840
    },
    {
      "epoch": 6.72,
      "learning_rate": 3.357487922705314e-05,
      "loss": 0.1288,
      "step": 2850
    },
    {
      "epoch": 6.75,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1382,
      "step": 2860
    },
    {
      "epoch": 6.77,
      "learning_rate": 3.3091787439613533e-05,
      "loss": 0.1276,
      "step": 2870
    },
    {
      "epoch": 6.79,
      "learning_rate": 3.2850241545893725e-05,
      "loss": 0.133,
      "step": 2880
    },
    {
      "epoch": 6.82,
      "learning_rate": 3.260869565217392e-05,
      "loss": 0.1336,
      "step": 2890
    },
    {
      "epoch": 6.84,
      "learning_rate": 3.236714975845411e-05,
      "loss": 0.1268,
      "step": 2900
    },
    {
      "epoch": 6.86,
      "learning_rate": 3.21256038647343e-05,
      "loss": 0.128,
      "step": 2910
    },
    {
      "epoch": 6.89,
      "learning_rate": 3.188405797101449e-05,
      "loss": 0.1253,
      "step": 2920
    },
    {
      "epoch": 6.91,
      "learning_rate": 3.164251207729469e-05,
      "loss": 0.1229,
      "step": 2930
    },
    {
      "epoch": 6.93,
      "learning_rate": 3.140096618357488e-05,
      "loss": 0.1314,
      "step": 2940
    },
    {
      "epoch": 6.96,
      "learning_rate": 3.1159420289855074e-05,
      "loss": 0.1269,
      "step": 2950
    },
    {
      "epoch": 6.98,
      "learning_rate": 3.0917874396135266e-05,
      "loss": 0.1257,
      "step": 2960
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.5197331309318542,
      "eval_runtime": 9.4923,
      "eval_samples_per_second": 79.327,
      "eval_steps_per_second": 5.057,
      "step": 2968
    },
    {
      "epoch": 7.0,
      "learning_rate": 3.067632850241546e-05,
      "loss": 0.1269,
      "step": 2970
    },
    {
      "epoch": 7.03,
      "learning_rate": 3.0434782608695656e-05,
      "loss": 0.1221,
      "step": 2980
    },
    {
      "epoch": 7.05,
      "learning_rate": 3.0193236714975848e-05,
      "loss": 0.1223,
      "step": 2990
    },
    {
      "epoch": 7.08,
      "learning_rate": 2.995169082125604e-05,
      "loss": 0.117,
      "step": 3000
    },
    {
      "epoch": 7.1,
      "learning_rate": 2.971014492753623e-05,
      "loss": 0.1182,
      "step": 3010
    },
    {
      "epoch": 7.12,
      "learning_rate": 2.9468599033816423e-05,
      "loss": 0.1276,
      "step": 3020
    },
    {
      "epoch": 7.15,
      "learning_rate": 2.9227053140096622e-05,
      "loss": 0.1224,
      "step": 3030
    },
    {
      "epoch": 7.17,
      "learning_rate": 2.8985507246376814e-05,
      "loss": 0.1282,
      "step": 3040
    },
    {
      "epoch": 7.19,
      "learning_rate": 2.8743961352657005e-05,
      "loss": 0.1151,
      "step": 3050
    },
    {
      "epoch": 7.22,
      "learning_rate": 2.8502415458937197e-05,
      "loss": 0.1227,
      "step": 3060
    },
    {
      "epoch": 7.24,
      "learning_rate": 2.826086956521739e-05,
      "loss": 0.1158,
      "step": 3070
    },
    {
      "epoch": 7.26,
      "learning_rate": 2.8019323671497587e-05,
      "loss": 0.1229,
      "step": 3080
    },
    {
      "epoch": 7.29,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1216,
      "step": 3090
    },
    {
      "epoch": 7.31,
      "learning_rate": 2.753623188405797e-05,
      "loss": 0.1206,
      "step": 3100
    },
    {
      "epoch": 7.33,
      "learning_rate": 2.7294685990338166e-05,
      "loss": 0.1228,
      "step": 3110
    },
    {
      "epoch": 7.36,
      "learning_rate": 2.7053140096618358e-05,
      "loss": 0.1247,
      "step": 3120
    },
    {
      "epoch": 7.38,
      "learning_rate": 2.6811594202898553e-05,
      "loss": 0.1211,
      "step": 3130
    },
    {
      "epoch": 7.41,
      "learning_rate": 2.6570048309178748e-05,
      "loss": 0.1232,
      "step": 3140
    },
    {
      "epoch": 7.43,
      "learning_rate": 2.632850241545894e-05,
      "loss": 0.1177,
      "step": 3150
    },
    {
      "epoch": 7.45,
      "learning_rate": 2.608695652173913e-05,
      "loss": 0.1242,
      "step": 3160
    },
    {
      "epoch": 7.48,
      "learning_rate": 2.5845410628019323e-05,
      "loss": 0.1164,
      "step": 3170
    },
    {
      "epoch": 7.5,
      "learning_rate": 2.5603864734299522e-05,
      "loss": 0.12,
      "step": 3180
    },
    {
      "epoch": 7.52,
      "learning_rate": 2.5362318840579714e-05,
      "loss": 0.1181,
      "step": 3190
    },
    {
      "epoch": 7.55,
      "learning_rate": 2.5120772946859905e-05,
      "loss": 0.1222,
      "step": 3200
    },
    {
      "epoch": 7.57,
      "learning_rate": 2.4879227053140097e-05,
      "loss": 0.1176,
      "step": 3210
    },
    {
      "epoch": 7.59,
      "learning_rate": 2.4637681159420292e-05,
      "loss": 0.1139,
      "step": 3220
    },
    {
      "epoch": 7.62,
      "learning_rate": 2.4396135265700484e-05,
      "loss": 0.1217,
      "step": 3230
    },
    {
      "epoch": 7.64,
      "learning_rate": 2.4154589371980676e-05,
      "loss": 0.1214,
      "step": 3240
    },
    {
      "epoch": 7.67,
      "learning_rate": 2.391304347826087e-05,
      "loss": 0.1153,
      "step": 3250
    },
    {
      "epoch": 7.69,
      "learning_rate": 2.3671497584541063e-05,
      "loss": 0.1242,
      "step": 3260
    },
    {
      "epoch": 7.71,
      "learning_rate": 2.3429951690821258e-05,
      "loss": 0.1207,
      "step": 3270
    },
    {
      "epoch": 7.74,
      "learning_rate": 2.318840579710145e-05,
      "loss": 0.1199,
      "step": 3280
    },
    {
      "epoch": 7.76,
      "learning_rate": 2.294685990338164e-05,
      "loss": 0.1225,
      "step": 3290
    },
    {
      "epoch": 7.78,
      "learning_rate": 2.2705314009661836e-05,
      "loss": 0.1206,
      "step": 3300
    },
    {
      "epoch": 7.81,
      "learning_rate": 2.246376811594203e-05,
      "loss": 0.1182,
      "step": 3310
    },
    {
      "epoch": 7.83,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1187,
      "step": 3320
    },
    {
      "epoch": 7.85,
      "learning_rate": 2.198067632850242e-05,
      "loss": 0.1167,
      "step": 3330
    },
    {
      "epoch": 7.88,
      "learning_rate": 2.173913043478261e-05,
      "loss": 0.1231,
      "step": 3340
    },
    {
      "epoch": 7.9,
      "learning_rate": 2.1497584541062805e-05,
      "loss": 0.1217,
      "step": 3350
    },
    {
      "epoch": 7.92,
      "learning_rate": 2.1256038647342997e-05,
      "loss": 0.1224,
      "step": 3360
    },
    {
      "epoch": 7.95,
      "learning_rate": 2.101449275362319e-05,
      "loss": 0.1153,
      "step": 3370
    },
    {
      "epoch": 7.97,
      "learning_rate": 2.0772946859903384e-05,
      "loss": 0.116,
      "step": 3380
    },
    {
      "epoch": 8.0,
      "learning_rate": 2.0531400966183576e-05,
      "loss": 0.1224,
      "step": 3390
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.5406513810157776,
      "eval_runtime": 9.543,
      "eval_samples_per_second": 78.906,
      "eval_steps_per_second": 5.03,
      "step": 3392
    },
    {
      "epoch": 8.02,
      "learning_rate": 2.028985507246377e-05,
      "loss": 0.1127,
      "step": 3400
    },
    {
      "epoch": 8.04,
      "learning_rate": 2.0048309178743963e-05,
      "loss": 0.1177,
      "step": 3410
    },
    {
      "epoch": 8.07,
      "learning_rate": 1.9806763285024154e-05,
      "loss": 0.1095,
      "step": 3420
    },
    {
      "epoch": 8.09,
      "learning_rate": 1.956521739130435e-05,
      "loss": 0.1117,
      "step": 3430
    },
    {
      "epoch": 8.11,
      "learning_rate": 1.932367149758454e-05,
      "loss": 0.1126,
      "step": 3440
    },
    {
      "epoch": 8.14,
      "learning_rate": 1.9082125603864733e-05,
      "loss": 0.1155,
      "step": 3450
    },
    {
      "epoch": 8.16,
      "learning_rate": 1.8840579710144928e-05,
      "loss": 0.1175,
      "step": 3460
    },
    {
      "epoch": 8.18,
      "learning_rate": 1.859903381642512e-05,
      "loss": 0.113,
      "step": 3470
    },
    {
      "epoch": 8.21,
      "learning_rate": 1.8357487922705315e-05,
      "loss": 0.1132,
      "step": 3480
    },
    {
      "epoch": 8.23,
      "learning_rate": 1.8115942028985507e-05,
      "loss": 0.1107,
      "step": 3490
    },
    {
      "epoch": 8.25,
      "learning_rate": 1.78743961352657e-05,
      "loss": 0.1114,
      "step": 3500
    },
    {
      "epoch": 8.28,
      "learning_rate": 1.7632850241545894e-05,
      "loss": 0.1128,
      "step": 3510
    },
    {
      "epoch": 8.3,
      "learning_rate": 1.739130434782609e-05,
      "loss": 0.113,
      "step": 3520
    },
    {
      "epoch": 8.33,
      "learning_rate": 1.714975845410628e-05,
      "loss": 0.1167,
      "step": 3530
    },
    {
      "epoch": 8.35,
      "learning_rate": 1.6908212560386476e-05,
      "loss": 0.1159,
      "step": 3540
    },
    {
      "epoch": 8.37,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.1132,
      "step": 3550
    },
    {
      "epoch": 8.4,
      "learning_rate": 1.6425120772946863e-05,
      "loss": 0.1138,
      "step": 3560
    },
    {
      "epoch": 8.42,
      "learning_rate": 1.6183574879227054e-05,
      "loss": 0.1171,
      "step": 3570
    },
    {
      "epoch": 8.44,
      "learning_rate": 1.5942028985507246e-05,
      "loss": 0.1139,
      "step": 3580
    },
    {
      "epoch": 8.47,
      "learning_rate": 1.570048309178744e-05,
      "loss": 0.118,
      "step": 3590
    },
    {
      "epoch": 8.49,
      "learning_rate": 1.5458937198067633e-05,
      "loss": 0.1163,
      "step": 3600
    },
    {
      "epoch": 8.51,
      "learning_rate": 1.5217391304347828e-05,
      "loss": 0.1191,
      "step": 3610
    },
    {
      "epoch": 8.54,
      "learning_rate": 1.497584541062802e-05,
      "loss": 0.1197,
      "step": 3620
    },
    {
      "epoch": 8.56,
      "learning_rate": 1.4734299516908212e-05,
      "loss": 0.123,
      "step": 3630
    },
    {
      "epoch": 8.58,
      "learning_rate": 1.4492753623188407e-05,
      "loss": 0.1157,
      "step": 3640
    },
    {
      "epoch": 8.61,
      "learning_rate": 1.4251207729468599e-05,
      "loss": 0.1048,
      "step": 3650
    },
    {
      "epoch": 8.63,
      "learning_rate": 1.4009661835748794e-05,
      "loss": 0.1119,
      "step": 3660
    },
    {
      "epoch": 8.66,
      "learning_rate": 1.3768115942028985e-05,
      "loss": 0.1136,
      "step": 3670
    },
    {
      "epoch": 8.68,
      "learning_rate": 1.3526570048309179e-05,
      "loss": 0.1176,
      "step": 3680
    },
    {
      "epoch": 8.7,
      "learning_rate": 1.3285024154589374e-05,
      "loss": 0.1238,
      "step": 3690
    },
    {
      "epoch": 8.73,
      "learning_rate": 1.3043478260869566e-05,
      "loss": 0.1098,
      "step": 3700
    },
    {
      "epoch": 8.75,
      "learning_rate": 1.2801932367149761e-05,
      "loss": 0.1052,
      "step": 3710
    },
    {
      "epoch": 8.77,
      "learning_rate": 1.2560386473429953e-05,
      "loss": 0.1145,
      "step": 3720
    },
    {
      "epoch": 8.8,
      "learning_rate": 1.2318840579710146e-05,
      "loss": 0.1097,
      "step": 3730
    },
    {
      "epoch": 8.82,
      "learning_rate": 1.2077294685990338e-05,
      "loss": 0.1095,
      "step": 3740
    },
    {
      "epoch": 8.84,
      "learning_rate": 1.1835748792270531e-05,
      "loss": 0.1175,
      "step": 3750
    },
    {
      "epoch": 8.87,
      "learning_rate": 1.1594202898550725e-05,
      "loss": 0.1176,
      "step": 3760
    },
    {
      "epoch": 8.89,
      "learning_rate": 1.1352657004830918e-05,
      "loss": 0.1169,
      "step": 3770
    },
    {
      "epoch": 8.92,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1073,
      "step": 3780
    },
    {
      "epoch": 8.94,
      "learning_rate": 1.0869565217391305e-05,
      "loss": 0.1143,
      "step": 3790
    },
    {
      "epoch": 8.96,
      "learning_rate": 1.0628019323671499e-05,
      "loss": 0.1195,
      "step": 3800
    },
    {
      "epoch": 8.99,
      "learning_rate": 1.0386473429951692e-05,
      "loss": 0.1191,
      "step": 3810
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.5499401092529297,
      "eval_runtime": 9.4806,
      "eval_samples_per_second": 79.426,
      "eval_steps_per_second": 5.063,
      "step": 3816
    },
    {
      "epoch": 9.01,
      "learning_rate": 1.0144927536231885e-05,
      "loss": 0.111,
      "step": 3820
    },
    {
      "epoch": 9.03,
      "learning_rate": 9.903381642512077e-06,
      "loss": 0.1113,
      "step": 3830
    },
    {
      "epoch": 9.06,
      "learning_rate": 9.66183574879227e-06,
      "loss": 0.1086,
      "step": 3840
    },
    {
      "epoch": 9.08,
      "learning_rate": 9.420289855072464e-06,
      "loss": 0.1089,
      "step": 3850
    },
    {
      "epoch": 9.1,
      "learning_rate": 9.178743961352658e-06,
      "loss": 0.1067,
      "step": 3860
    },
    {
      "epoch": 9.13,
      "learning_rate": 8.93719806763285e-06,
      "loss": 0.109,
      "step": 3870
    },
    {
      "epoch": 9.15,
      "learning_rate": 8.695652173913044e-06,
      "loss": 0.1111,
      "step": 3880
    },
    {
      "epoch": 9.17,
      "learning_rate": 8.454106280193238e-06,
      "loss": 0.1126,
      "step": 3890
    },
    {
      "epoch": 9.2,
      "learning_rate": 8.212560386473431e-06,
      "loss": 0.1105,
      "step": 3900
    },
    {
      "epoch": 9.22,
      "learning_rate": 7.971014492753623e-06,
      "loss": 0.1093,
      "step": 3910
    },
    {
      "epoch": 9.25,
      "learning_rate": 7.729468599033817e-06,
      "loss": 0.109,
      "step": 3920
    },
    {
      "epoch": 9.27,
      "learning_rate": 7.48792270531401e-06,
      "loss": 0.1155,
      "step": 3930
    },
    {
      "epoch": 9.29,
      "learning_rate": 7.246376811594203e-06,
      "loss": 0.1119,
      "step": 3940
    },
    {
      "epoch": 9.32,
      "learning_rate": 7.004830917874397e-06,
      "loss": 0.1088,
      "step": 3950
    },
    {
      "epoch": 9.34,
      "learning_rate": 6.7632850241545894e-06,
      "loss": 0.1097,
      "step": 3960
    },
    {
      "epoch": 9.36,
      "learning_rate": 6.521739130434783e-06,
      "loss": 0.1126,
      "step": 3970
    },
    {
      "epoch": 9.39,
      "learning_rate": 6.280193236714976e-06,
      "loss": 0.105,
      "step": 3980
    },
    {
      "epoch": 9.41,
      "learning_rate": 6.038647342995169e-06,
      "loss": 0.109,
      "step": 3990
    },
    {
      "epoch": 9.43,
      "learning_rate": 5.797101449275362e-06,
      "loss": 0.1093,
      "step": 4000
    }
  ],
  "max_steps": 4240,
  "num_train_epochs": 10,
  "total_flos": 0.0,
  "trial_name": null,
  "trial_params": null
}