{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 20.0,
  "eval_steps": 500,
  "global_step": 372000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05,
      "learning_rate": 3.125e-06,
      "loss": 7.545,
      "step": 1000
    },
    {
      "epoch": 0.11,
      "learning_rate": 6.25e-06,
      "loss": 5.8075,
      "step": 2000
    },
    {
      "epoch": 0.16,
      "learning_rate": 9.375000000000001e-06,
      "loss": 5.3765,
      "step": 3000
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.25e-05,
      "loss": 5.1503,
      "step": 4000
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.5625e-05,
      "loss": 5.0086,
      "step": 5000
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 4.8629,
      "step": 6000
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.1875e-05,
      "loss": 4.7453,
      "step": 7000
    },
    {
      "epoch": 0.43,
      "learning_rate": 2.5e-05,
      "loss": 4.6427,
      "step": 8000
    },
    {
      "epoch": 0.48,
      "learning_rate": 2.8125000000000003e-05,
      "loss": 4.5521,
      "step": 9000
    },
    {
      "epoch": 0.54,
      "learning_rate": 3.125e-05,
      "loss": 4.4683,
      "step": 10000
    },
    {
      "epoch": 0.59,
      "learning_rate": 3.4371875e-05,
      "loss": 4.4037,
      "step": 11000
    },
    {
      "epoch": 0.65,
      "learning_rate": 3.7496875e-05,
      "loss": 4.3415,
      "step": 12000
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.0621875e-05,
      "loss": 4.2805,
      "step": 13000
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.374375e-05,
      "loss": 4.2289,
      "step": 14000
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.686875e-05,
      "loss": 4.1822,
      "step": 15000
    },
    {
      "epoch": 0.86,
      "learning_rate": 4.999375e-05,
      "loss": 4.1383,
      "step": 16000
    },
    {
      "epoch": 0.91,
      "learning_rate": 5.311875000000001e-05,
      "loss": 4.0872,
      "step": 17000
    },
    {
      "epoch": 0.97,
      "learning_rate": 5.6240625e-05,
      "loss": 4.0521,
      "step": 18000
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.3096176628624153,
      "eval_loss": 4.2758893966674805,
      "eval_runtime": 152.539,
      "eval_samples_per_second": 379.693,
      "eval_steps_per_second": 5.933,
      "step": 18600
    },
    {
      "epoch": 1.02,
      "learning_rate": 5.93625e-05,
      "loss": 4.0001,
      "step": 19000
    },
    {
      "epoch": 1.08,
      "learning_rate": 6.24875e-05,
      "loss": 3.9629,
      "step": 20000
    },
    {
      "epoch": 1.13,
      "learning_rate": 6.560937500000001e-05,
      "loss": 3.9269,
      "step": 21000
    },
    {
      "epoch": 1.18,
      "learning_rate": 6.873125000000001e-05,
      "loss": 3.8872,
      "step": 22000
    },
    {
      "epoch": 1.24,
      "learning_rate": 7.185625e-05,
      "loss": 3.8505,
      "step": 23000
    },
    {
      "epoch": 1.29,
      "learning_rate": 7.4978125e-05,
      "loss": 3.8219,
      "step": 24000
    },
    {
      "epoch": 1.34,
      "learning_rate": 7.81e-05,
      "loss": 3.7983,
      "step": 25000
    },
    {
      "epoch": 1.4,
      "learning_rate": 8.122500000000001e-05,
      "loss": 3.769,
      "step": 26000
    },
    {
      "epoch": 1.45,
      "learning_rate": 8.434687500000001e-05,
      "loss": 3.7446,
      "step": 27000
    },
    {
      "epoch": 1.51,
      "learning_rate": 8.7471875e-05,
      "loss": 3.7254,
      "step": 28000
    },
    {
      "epoch": 1.56,
      "learning_rate": 9.0596875e-05,
      "loss": 3.7042,
      "step": 29000
    },
    {
      "epoch": 1.61,
      "learning_rate": 9.371875e-05,
      "loss": 3.688,
      "step": 30000
    },
    {
      "epoch": 1.67,
      "learning_rate": 9.684375e-05,
      "loss": 3.6685,
      "step": 31000
    },
    {
      "epoch": 1.72,
      "learning_rate": 9.9965625e-05,
      "loss": 3.6542,
      "step": 32000
    },
    {
      "epoch": 1.77,
      "learning_rate": 9.970911764705882e-05,
      "loss": 3.6318,
      "step": 33000
    },
    {
      "epoch": 1.83,
      "learning_rate": 9.941529411764706e-05,
      "loss": 3.6132,
      "step": 34000
    },
    {
      "epoch": 1.88,
      "learning_rate": 9.912117647058824e-05,
      "loss": 3.5971,
      "step": 35000
    },
    {
      "epoch": 1.94,
      "learning_rate": 9.882705882352942e-05,
      "loss": 3.5801,
      "step": 36000
    },
    {
      "epoch": 1.99,
      "learning_rate": 9.853294117647059e-05,
      "loss": 3.567,
      "step": 37000
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3623238704713749,
      "eval_loss": 3.751614570617676,
      "eval_runtime": 154.1657,
      "eval_samples_per_second": 375.687,
      "eval_steps_per_second": 5.87,
      "step": 37200
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.823882352941177e-05,
      "loss": 3.5234,
      "step": 38000
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.794500000000001e-05,
      "loss": 3.5104,
      "step": 39000
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.765117647058824e-05,
      "loss": 3.5032,
      "step": 40000
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.735705882352942e-05,
      "loss": 3.4944,
      "step": 41000
    },
    {
      "epoch": 2.26,
      "learning_rate": 9.70629411764706e-05,
      "loss": 3.4852,
      "step": 42000
    },
    {
      "epoch": 2.31,
      "learning_rate": 9.676882352941177e-05,
      "loss": 3.4756,
      "step": 43000
    },
    {
      "epoch": 2.37,
      "learning_rate": 9.647470588235295e-05,
      "loss": 3.4617,
      "step": 44000
    },
    {
      "epoch": 2.42,
      "learning_rate": 9.618088235294118e-05,
      "loss": 3.4569,
      "step": 45000
    },
    {
      "epoch": 2.47,
      "learning_rate": 9.588676470588235e-05,
      "loss": 3.4534,
      "step": 46000
    },
    {
      "epoch": 2.53,
      "learning_rate": 9.559323529411765e-05,
      "loss": 3.4463,
      "step": 47000
    },
    {
      "epoch": 2.58,
      "learning_rate": 9.529911764705884e-05,
      "loss": 3.4381,
      "step": 48000
    },
    {
      "epoch": 2.63,
      "learning_rate": 9.5005e-05,
      "loss": 3.4265,
      "step": 49000
    },
    {
      "epoch": 2.69,
      "learning_rate": 9.471088235294118e-05,
      "loss": 3.425,
      "step": 50000
    },
    {
      "epoch": 2.74,
      "learning_rate": 9.441705882352941e-05,
      "loss": 3.4179,
      "step": 51000
    },
    {
      "epoch": 2.8,
      "learning_rate": 9.412294117647059e-05,
      "loss": 3.4093,
      "step": 52000
    },
    {
      "epoch": 2.85,
      "learning_rate": 9.382911764705883e-05,
      "loss": 3.4023,
      "step": 53000
    },
    {
      "epoch": 2.9,
      "learning_rate": 9.3535e-05,
      "loss": 3.396,
      "step": 54000
    },
    {
      "epoch": 2.96,
      "learning_rate": 9.324117647058825e-05,
      "loss": 3.3864,
      "step": 55000
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.38018321354100115,
      "eval_loss": 3.5931003093719482,
      "eval_runtime": 154.2085,
      "eval_samples_per_second": 375.582,
      "eval_steps_per_second": 5.869,
      "step": 55800
    },
    {
      "epoch": 3.01,
      "learning_rate": 9.294705882352942e-05,
      "loss": 3.3746,
      "step": 56000
    },
    {
      "epoch": 3.06,
      "learning_rate": 9.26529411764706e-05,
      "loss": 3.3329,
      "step": 57000
    },
    {
      "epoch": 3.12,
      "learning_rate": 9.235882352941176e-05,
      "loss": 3.3298,
      "step": 58000
    },
    {
      "epoch": 3.17,
      "learning_rate": 9.206500000000001e-05,
      "loss": 3.326,
      "step": 59000
    },
    {
      "epoch": 3.23,
      "learning_rate": 9.177117647058824e-05,
      "loss": 3.3312,
      "step": 60000
    },
    {
      "epoch": 3.28,
      "learning_rate": 9.147705882352941e-05,
      "loss": 3.3201,
      "step": 61000
    },
    {
      "epoch": 3.33,
      "learning_rate": 9.118294117647059e-05,
      "loss": 3.3252,
      "step": 62000
    },
    {
      "epoch": 3.39,
      "learning_rate": 9.088882352941177e-05,
      "loss": 3.3186,
      "step": 63000
    },
    {
      "epoch": 3.44,
      "learning_rate": 9.059500000000001e-05,
      "loss": 3.3134,
      "step": 64000
    },
    {
      "epoch": 3.49,
      "learning_rate": 9.030117647058824e-05,
      "loss": 3.3148,
      "step": 65000
    },
    {
      "epoch": 3.55,
      "learning_rate": 9.000705882352942e-05,
      "loss": 3.3092,
      "step": 66000
    },
    {
      "epoch": 3.6,
      "learning_rate": 8.971294117647059e-05,
      "loss": 3.3104,
      "step": 67000
    },
    {
      "epoch": 3.66,
      "learning_rate": 8.941882352941177e-05,
      "loss": 3.3048,
      "step": 68000
    },
    {
      "epoch": 3.71,
      "learning_rate": 8.9125e-05,
      "loss": 3.3012,
      "step": 69000
    },
    {
      "epoch": 3.76,
      "learning_rate": 8.883088235294117e-05,
      "loss": 3.2979,
      "step": 70000
    },
    {
      "epoch": 3.82,
      "learning_rate": 8.853735294117647e-05,
      "loss": 3.2973,
      "step": 71000
    },
    {
      "epoch": 3.87,
      "learning_rate": 8.824323529411766e-05,
      "loss": 3.2955,
      "step": 72000
    },
    {
      "epoch": 3.92,
      "learning_rate": 8.794911764705884e-05,
      "loss": 3.2902,
      "step": 73000
    },
    {
      "epoch": 3.98,
      "learning_rate": 8.765500000000001e-05,
      "loss": 3.2901,
      "step": 74000
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3882912820661655,
      "eval_loss": 3.5232033729553223,
      "eval_runtime": 154.1118,
      "eval_samples_per_second": 375.818,
      "eval_steps_per_second": 5.872,
      "step": 74400
    },
    {
      "epoch": 4.03,
      "learning_rate": 8.73614705882353e-05,
      "loss": 3.2518,
      "step": 75000
    },
    {
      "epoch": 4.09,
      "learning_rate": 8.706735294117647e-05,
      "loss": 3.2301,
      "step": 76000
    },
    {
      "epoch": 4.14,
      "learning_rate": 8.67735294117647e-05,
      "loss": 3.2344,
      "step": 77000
    },
    {
      "epoch": 4.19,
      "learning_rate": 8.647941176470589e-05,
      "loss": 3.2324,
      "step": 78000
    },
    {
      "epoch": 4.25,
      "learning_rate": 8.618529411764707e-05,
      "loss": 3.2311,
      "step": 79000
    },
    {
      "epoch": 4.3,
      "learning_rate": 8.58914705882353e-05,
      "loss": 3.2382,
      "step": 80000
    },
    {
      "epoch": 4.35,
      "learning_rate": 8.559735294117648e-05,
      "loss": 3.2331,
      "step": 81000
    },
    {
      "epoch": 4.41,
      "learning_rate": 8.530352941176472e-05,
      "loss": 3.2297,
      "step": 82000
    },
    {
      "epoch": 4.46,
      "learning_rate": 8.500941176470588e-05,
      "loss": 3.2285,
      "step": 83000
    },
    {
      "epoch": 4.52,
      "learning_rate": 8.471529411764706e-05,
      "loss": 3.2306,
      "step": 84000
    },
    {
      "epoch": 4.57,
      "learning_rate": 8.442117647058823e-05,
      "loss": 3.2263,
      "step": 85000
    },
    {
      "epoch": 4.62,
      "learning_rate": 8.412735294117648e-05,
      "loss": 3.2272,
      "step": 86000
    },
    {
      "epoch": 4.68,
      "learning_rate": 8.383323529411765e-05,
      "loss": 3.2249,
      "step": 87000
    },
    {
      "epoch": 4.73,
      "learning_rate": 8.353911764705883e-05,
      "loss": 3.23,
      "step": 88000
    },
    {
      "epoch": 4.78,
      "learning_rate": 8.324500000000001e-05,
      "loss": 3.2313,
      "step": 89000
    },
    {
      "epoch": 4.84,
      "learning_rate": 8.295117647058825e-05,
      "loss": 3.2227,
      "step": 90000
    },
    {
      "epoch": 4.89,
      "learning_rate": 8.265705882352943e-05,
      "loss": 3.2254,
      "step": 91000
    },
    {
      "epoch": 4.95,
      "learning_rate": 8.236323529411764e-05,
      "loss": 3.2206,
      "step": 92000
    },
    {
      "epoch": 5.0,
      "learning_rate": 8.206911764705882e-05,
      "loss": 3.2176,
      "step": 93000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3938886898060494,
      "eval_loss": 3.4593958854675293,
      "eval_runtime": 153.8963,
      "eval_samples_per_second": 376.344,
      "eval_steps_per_second": 5.881,
      "step": 93000
    },
    {
      "epoch": 5.05,
      "learning_rate": 8.1775e-05,
      "loss": 3.1614,
      "step": 94000
    },
    {
      "epoch": 5.11,
      "learning_rate": 8.148088235294119e-05,
      "loss": 3.1639,
      "step": 95000
    },
    {
      "epoch": 5.16,
      "learning_rate": 8.118705882352942e-05,
      "loss": 3.1664,
      "step": 96000
    },
    {
      "epoch": 5.22,
      "learning_rate": 8.089323529411766e-05,
      "loss": 3.1726,
      "step": 97000
    },
    {
      "epoch": 5.27,
      "learning_rate": 8.059911764705883e-05,
      "loss": 3.1677,
      "step": 98000
    },
    {
      "epoch": 5.32,
      "learning_rate": 8.030529411764706e-05,
      "loss": 3.1686,
      "step": 99000
    },
    {
      "epoch": 5.38,
      "learning_rate": 8.001117647058823e-05,
      "loss": 3.1734,
      "step": 100000
    },
    {
      "epoch": 5.43,
      "learning_rate": 7.971705882352942e-05,
      "loss": 3.1742,
      "step": 101000
    },
    {
      "epoch": 5.48,
      "learning_rate": 7.94229411764706e-05,
      "loss": 3.175,
      "step": 102000
    },
    {
      "epoch": 5.54,
      "learning_rate": 7.912911764705882e-05,
      "loss": 3.1693,
      "step": 103000
    },
    {
      "epoch": 5.59,
      "learning_rate": 7.8835e-05,
      "loss": 3.1727,
      "step": 104000
    },
    {
      "epoch": 5.65,
      "learning_rate": 7.854117647058824e-05,
      "loss": 3.1715,
      "step": 105000
    },
    {
      "epoch": 5.7,
      "learning_rate": 7.824705882352942e-05,
      "loss": 3.1722,
      "step": 106000
    },
    {
      "epoch": 5.75,
      "learning_rate": 7.79529411764706e-05,
      "loss": 3.1689,
      "step": 107000
    },
    {
      "epoch": 5.81,
      "learning_rate": 7.765941176470588e-05,
      "loss": 3.1635,
      "step": 108000
    },
    {
      "epoch": 5.86,
      "learning_rate": 7.736529411764706e-05,
      "loss": 3.1679,
      "step": 109000
    },
    {
      "epoch": 5.91,
      "learning_rate": 7.707117647058823e-05,
      "loss": 3.1717,
      "step": 110000
    },
    {
      "epoch": 5.97,
      "learning_rate": 7.677705882352941e-05,
      "loss": 3.1641,
      "step": 111000
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.3960908505692269,
      "eval_loss": 3.4612395763397217,
      "eval_runtime": 154.0615,
      "eval_samples_per_second": 375.941,
      "eval_steps_per_second": 5.874,
      "step": 111600
    },
    {
      "epoch": 6.02,
      "learning_rate": 7.648323529411765e-05,
      "loss": 3.1442,
      "step": 112000
    },
    {
      "epoch": 6.08,
      "learning_rate": 7.618941176470588e-05,
      "loss": 3.1105,
      "step": 113000
    },
    {
      "epoch": 6.13,
      "learning_rate": 7.589529411764707e-05,
      "loss": 3.1109,
      "step": 114000
    },
    {
      "epoch": 6.18,
      "learning_rate": 7.560117647058825e-05,
      "loss": 3.1198,
      "step": 115000
    },
    {
      "epoch": 6.24,
      "learning_rate": 7.530705882352942e-05,
      "loss": 3.1211,
      "step": 116000
    },
    {
      "epoch": 6.29,
      "learning_rate": 7.501323529411764e-05,
      "loss": 3.1237,
      "step": 117000
    },
    {
      "epoch": 6.34,
      "learning_rate": 7.471911764705882e-05,
      "loss": 3.1238,
      "step": 118000
    },
    {
      "epoch": 6.4,
      "learning_rate": 7.4425e-05,
      "loss": 3.1248,
      "step": 119000
    },
    {
      "epoch": 6.45,
      "learning_rate": 7.413088235294118e-05,
      "loss": 3.1232,
      "step": 120000
    },
    {
      "epoch": 6.51,
      "learning_rate": 7.383705882352941e-05,
      "loss": 3.1217,
      "step": 121000
    },
    {
      "epoch": 6.56,
      "learning_rate": 7.354294117647059e-05,
      "loss": 3.1193,
      "step": 122000
    },
    {
      "epoch": 6.61,
      "learning_rate": 7.324882352941177e-05,
      "loss": 3.1205,
      "step": 123000
    },
    {
      "epoch": 6.67,
      "learning_rate": 7.295500000000001e-05,
      "loss": 3.1316,
      "step": 124000
    },
    {
      "epoch": 6.72,
      "learning_rate": 7.266088235294119e-05,
      "loss": 3.1248,
      "step": 125000
    },
    {
      "epoch": 6.77,
      "learning_rate": 7.236705882352941e-05,
      "loss": 3.1276,
      "step": 126000
    },
    {
      "epoch": 6.83,
      "learning_rate": 7.207294117647059e-05,
      "loss": 3.127,
      "step": 127000
    },
    {
      "epoch": 6.88,
      "learning_rate": 7.177911764705882e-05,
      "loss": 3.1266,
      "step": 128000
    },
    {
      "epoch": 6.94,
      "learning_rate": 7.1485e-05,
      "loss": 3.128,
      "step": 129000
    },
    {
      "epoch": 6.99,
      "learning_rate": 7.119117647058824e-05,
      "loss": 3.1229,
      "step": 130000
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4000027947737194,
      "eval_loss": 3.415517568588257,
      "eval_runtime": 154.2934,
      "eval_samples_per_second": 375.376,
      "eval_steps_per_second": 5.865,
      "step": 130200
    },
    {
      "epoch": 7.04,
      "learning_rate": 7.089705882352942e-05,
      "loss": 3.0796,
      "step": 131000
    },
    {
      "epoch": 7.1,
      "learning_rate": 7.060323529411765e-05,
      "loss": 3.074,
      "step": 132000
    },
    {
      "epoch": 7.15,
      "learning_rate": 7.030911764705884e-05,
      "loss": 3.0722,
      "step": 133000
    },
    {
      "epoch": 7.2,
      "learning_rate": 7.001500000000001e-05,
      "loss": 3.0796,
      "step": 134000
    },
    {
      "epoch": 7.26,
      "learning_rate": 6.972088235294118e-05,
      "loss": 3.0844,
      "step": 135000
    },
    {
      "epoch": 7.31,
      "learning_rate": 6.94270588235294e-05,
      "loss": 3.0789,
      "step": 136000
    },
    {
      "epoch": 7.37,
      "learning_rate": 6.91329411764706e-05,
      "loss": 3.0785,
      "step": 137000
    },
    {
      "epoch": 7.42,
      "learning_rate": 6.883882352941177e-05,
      "loss": 3.0841,
      "step": 138000
    },
    {
      "epoch": 7.47,
      "learning_rate": 6.8545e-05,
      "loss": 3.0857,
      "step": 139000
    },
    {
      "epoch": 7.53,
      "learning_rate": 6.825088235294118e-05,
      "loss": 3.0871,
      "step": 140000
    },
    {
      "epoch": 7.58,
      "learning_rate": 6.795705882352942e-05,
      "loss": 3.0882,
      "step": 141000
    },
    {
      "epoch": 7.63,
      "learning_rate": 6.76629411764706e-05,
      "loss": 3.0911,
      "step": 142000
    },
    {
      "epoch": 7.69,
      "learning_rate": 6.736911764705883e-05,
      "loss": 3.0885,
      "step": 143000
    },
    {
      "epoch": 7.74,
      "learning_rate": 6.707529411764706e-05,
      "loss": 3.0894,
      "step": 144000
    },
    {
      "epoch": 7.8,
      "learning_rate": 6.678117647058823e-05,
      "loss": 3.0877,
      "step": 145000
    },
    {
      "epoch": 7.85,
      "learning_rate": 6.648735294117648e-05,
      "loss": 3.0882,
      "step": 146000
    },
    {
      "epoch": 7.9,
      "learning_rate": 6.619323529411765e-05,
      "loss": 3.0921,
      "step": 147000
    },
    {
      "epoch": 7.96,
      "learning_rate": 6.589911764705883e-05,
      "loss": 3.0932,
      "step": 148000
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.40145466628453513,
      "eval_loss": 3.406418561935425,
      "eval_runtime": 153.771,
      "eval_samples_per_second": 376.651,
      "eval_steps_per_second": 5.885,
      "step": 148800
    },
    {
      "epoch": 8.01,
      "learning_rate": 6.5605e-05,
      "loss": 3.0811,
      "step": 149000
    },
    {
      "epoch": 8.06,
      "learning_rate": 6.531088235294118e-05,
      "loss": 3.035,
      "step": 150000
    },
    {
      "epoch": 8.12,
      "learning_rate": 6.501676470588236e-05,
      "loss": 3.0444,
      "step": 151000
    },
    {
      "epoch": 8.17,
      "learning_rate": 6.472323529411764e-05,
      "loss": 3.0449,
      "step": 152000
    },
    {
      "epoch": 8.23,
      "learning_rate": 6.442911764705882e-05,
      "loss": 3.0472,
      "step": 153000
    },
    {
      "epoch": 8.28,
      "learning_rate": 6.4135e-05,
      "loss": 3.0433,
      "step": 154000
    },
    {
      "epoch": 8.33,
      "learning_rate": 6.384088235294117e-05,
      "loss": 3.0508,
      "step": 155000
    },
    {
      "epoch": 8.39,
      "learning_rate": 6.354676470588236e-05,
      "loss": 3.0504,
      "step": 156000
    },
    {
      "epoch": 8.44,
      "learning_rate": 6.325294117647059e-05,
      "loss": 3.0502,
      "step": 157000
    },
    {
      "epoch": 8.49,
      "learning_rate": 6.295882352941177e-05,
      "loss": 3.0529,
      "step": 158000
    },
    {
      "epoch": 8.55,
      "learning_rate": 6.266500000000001e-05,
      "loss": 3.0556,
      "step": 159000
    },
    {
      "epoch": 8.6,
      "learning_rate": 6.237088235294119e-05,
      "loss": 3.0534,
      "step": 160000
    },
    {
      "epoch": 8.66,
      "learning_rate": 6.207676470588235e-05,
      "loss": 3.055,
      "step": 161000
    },
    {
      "epoch": 8.71,
      "learning_rate": 6.178264705882353e-05,
      "loss": 3.0539,
      "step": 162000
    },
    {
      "epoch": 8.76,
      "learning_rate": 6.14885294117647e-05,
      "loss": 3.0555,
      "step": 163000
    },
    {
      "epoch": 8.82,
      "learning_rate": 6.119470588235295e-05,
      "loss": 3.0564,
      "step": 164000
    },
    {
      "epoch": 8.87,
      "learning_rate": 6.090058823529412e-05,
      "loss": 3.0583,
      "step": 165000
    },
    {
      "epoch": 8.92,
      "learning_rate": 6.06064705882353e-05,
      "loss": 3.0592,
      "step": 166000
    },
    {
      "epoch": 8.98,
      "learning_rate": 6.0312647058823536e-05,
      "loss": 3.0577,
      "step": 167000
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.4035858827917586,
      "eval_loss": 3.4074039459228516,
      "eval_runtime": 153.9331,
      "eval_samples_per_second": 376.254,
      "eval_steps_per_second": 5.879,
      "step": 167400
    },
    {
      "epoch": 9.03,
      "learning_rate": 6.001852941176471e-05,
      "loss": 3.0296,
      "step": 168000
    },
    {
      "epoch": 9.09,
      "learning_rate": 5.972470588235295e-05,
      "loss": 3.0068,
      "step": 169000
    },
    {
      "epoch": 9.14,
      "learning_rate": 5.943058823529412e-05,
      "loss": 3.0102,
      "step": 170000
    },
    {
      "epoch": 9.19,
      "learning_rate": 5.9136470588235295e-05,
      "loss": 3.0193,
      "step": 171000
    },
    {
      "epoch": 9.25,
      "learning_rate": 5.884235294117647e-05,
      "loss": 3.012,
      "step": 172000
    },
    {
      "epoch": 9.3,
      "learning_rate": 5.854882352941177e-05,
      "loss": 3.0217,
      "step": 173000
    },
    {
      "epoch": 9.35,
      "learning_rate": 5.8254705882352944e-05,
      "loss": 3.0227,
      "step": 174000
    },
    {
      "epoch": 9.41,
      "learning_rate": 5.796058823529412e-05,
      "loss": 3.0236,
      "step": 175000
    },
    {
      "epoch": 9.46,
      "learning_rate": 5.76664705882353e-05,
      "loss": 3.0226,
      "step": 176000
    },
    {
      "epoch": 9.52,
      "learning_rate": 5.737235294117648e-05,
      "loss": 3.0263,
      "step": 177000
    },
    {
      "epoch": 9.57,
      "learning_rate": 5.707823529411766e-05,
      "loss": 3.0265,
      "step": 178000
    },
    {
      "epoch": 9.62,
      "learning_rate": 5.678441176470588e-05,
      "loss": 3.0291,
      "step": 179000
    },
    {
      "epoch": 9.68,
      "learning_rate": 5.6490588235294116e-05,
      "loss": 3.0286,
      "step": 180000
    },
    {
      "epoch": 9.73,
      "learning_rate": 5.619676470588235e-05,
      "loss": 3.0275,
      "step": 181000
    },
    {
      "epoch": 9.78,
      "learning_rate": 5.590264705882353e-05,
      "loss": 3.0297,
      "step": 182000
    },
    {
      "epoch": 9.84,
      "learning_rate": 5.560852941176471e-05,
      "loss": 3.0305,
      "step": 183000
    },
    {
      "epoch": 9.89,
      "learning_rate": 5.531441176470589e-05,
      "loss": 3.0287,
      "step": 184000
    },
    {
      "epoch": 9.95,
      "learning_rate": 5.5020294117647065e-05,
      "loss": 3.0281,
      "step": 185000
    },
    {
      "epoch": 10.0,
      "learning_rate": 5.47264705882353e-05,
      "loss": 3.0285,
      "step": 186000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.40581686465891736,
      "eval_loss": 3.394498109817505,
      "eval_runtime": 153.8811,
      "eval_samples_per_second": 376.382,
      "eval_steps_per_second": 5.881,
      "step": 186000
    },
    {
      "epoch": 10.05,
      "learning_rate": 5.4432352941176464e-05,
      "loss": 2.9757,
      "step": 187000
    },
    {
      "epoch": 10.11,
      "learning_rate": 5.413882352941177e-05,
      "loss": 2.9815,
      "step": 188000
    },
    {
      "epoch": 10.16,
      "learning_rate": 5.3844705882352944e-05,
      "loss": 2.9846,
      "step": 189000
    },
    {
      "epoch": 10.22,
      "learning_rate": 5.355058823529412e-05,
      "loss": 2.9881,
      "step": 190000
    },
    {
      "epoch": 10.27,
      "learning_rate": 5.325676470588236e-05,
      "loss": 2.9886,
      "step": 191000
    },
    {
      "epoch": 10.32,
      "learning_rate": 5.296264705882353e-05,
      "loss": 2.9967,
      "step": 192000
    },
    {
      "epoch": 10.38,
      "learning_rate": 5.266852941176471e-05,
      "loss": 2.9988,
      "step": 193000
    },
    {
      "epoch": 10.43,
      "learning_rate": 5.2374411764705886e-05,
      "loss": 2.9983,
      "step": 194000
    },
    {
      "epoch": 10.48,
      "learning_rate": 5.208058823529412e-05,
      "loss": 2.9995,
      "step": 195000
    },
    {
      "epoch": 10.54,
      "learning_rate": 5.178647058823529e-05,
      "loss": 2.9981,
      "step": 196000
    },
    {
      "epoch": 10.59,
      "learning_rate": 5.149235294117647e-05,
      "loss": 2.9981,
      "step": 197000
    },
    {
      "epoch": 10.65,
      "learning_rate": 5.1198529411764705e-05,
      "loss": 3.0073,
      "step": 198000
    },
    {
      "epoch": 10.7,
      "learning_rate": 5.090441176470588e-05,
      "loss": 3.0073,
      "step": 199000
    },
    {
      "epoch": 10.75,
      "learning_rate": 5.061058823529412e-05,
      "loss": 3.0049,
      "step": 200000
    },
    {
      "epoch": 10.81,
      "learning_rate": 5.03164705882353e-05,
      "loss": 3.0007,
      "step": 201000
    },
    {
      "epoch": 10.86,
      "learning_rate": 5.002235294117648e-05,
      "loss": 3.0053,
      "step": 202000
    },
    {
      "epoch": 10.91,
      "learning_rate": 4.972823529411765e-05,
      "loss": 3.0053,
      "step": 203000
    },
    {
      "epoch": 10.97,
      "learning_rate": 4.9434411764705884e-05,
      "loss": 3.0042,
      "step": 204000
    },
    {
      "epoch": 11.0,
      "eval_accuracy": 0.4052487059727405,
      "eval_loss": 3.396245241165161,
      "eval_runtime": 154.4165,
      "eval_samples_per_second": 375.077,
      "eval_steps_per_second": 5.861,
      "step": 204600
    },
    {
      "epoch": 11.02,
      "learning_rate": 4.914029411764706e-05,
      "loss": 2.9824,
      "step": 205000
    },
    {
      "epoch": 11.08,
      "learning_rate": 4.884617647058824e-05,
      "loss": 2.956,
      "step": 206000
    },
    {
      "epoch": 11.13,
      "learning_rate": 4.855235294117647e-05,
      "loss": 2.9598,
      "step": 207000
    },
    {
      "epoch": 11.18,
      "learning_rate": 4.825823529411765e-05,
      "loss": 2.9617,
      "step": 208000
    },
    {
      "epoch": 11.24,
      "learning_rate": 4.7964117647058826e-05,
      "loss": 2.9681,
      "step": 209000
    },
    {
      "epoch": 11.29,
      "learning_rate": 4.767029411764706e-05,
      "loss": 2.9665,
      "step": 210000
    },
    {
      "epoch": 11.34,
      "learning_rate": 4.737617647058823e-05,
      "loss": 2.9715,
      "step": 211000
    },
    {
      "epoch": 11.4,
      "learning_rate": 4.7082058823529416e-05,
      "loss": 2.9709,
      "step": 212000
    },
    {
      "epoch": 11.45,
      "learning_rate": 4.678823529411765e-05,
      "loss": 2.9727,
      "step": 213000
    },
    {
      "epoch": 11.51,
      "learning_rate": 4.649441176470589e-05,
      "loss": 2.9735,
      "step": 214000
    },
    {
      "epoch": 11.56,
      "learning_rate": 4.620029411764706e-05,
      "loss": 2.979,
      "step": 215000
    },
    {
      "epoch": 11.61,
      "learning_rate": 4.5906176470588234e-05,
      "loss": 2.9775,
      "step": 216000
    },
    {
      "epoch": 11.67,
      "learning_rate": 4.561235294117647e-05,
      "loss": 2.9775,
      "step": 217000
    },
    {
      "epoch": 11.72,
      "learning_rate": 4.531823529411765e-05,
      "loss": 2.9818,
      "step": 218000
    },
    {
      "epoch": 11.77,
      "learning_rate": 4.502411764705883e-05,
      "loss": 2.982,
      "step": 219000
    },
    {
      "epoch": 11.83,
      "learning_rate": 4.473e-05,
      "loss": 2.9825,
      "step": 220000
    },
    {
      "epoch": 11.88,
      "learning_rate": 4.443588235294118e-05,
      "loss": 2.9857,
      "step": 221000
    },
    {
      "epoch": 11.94,
      "learning_rate": 4.414205882352941e-05,
      "loss": 2.9834,
      "step": 222000
    },
    {
      "epoch": 11.99,
      "learning_rate": 4.384794117647059e-05,
      "loss": 2.9833,
      "step": 223000
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.4059758845962687,
      "eval_loss": 3.3878443241119385,
      "eval_runtime": 154.1343,
      "eval_samples_per_second": 375.763,
      "eval_steps_per_second": 5.872,
      "step": 223200
    },
    {
      "epoch": 12.04,
      "learning_rate": 4.3553823529411766e-05,
      "loss": 2.9442,
      "step": 224000
    },
    {
      "epoch": 12.1,
      "learning_rate": 4.326029411764706e-05,
      "loss": 2.9357,
      "step": 225000
    },
    {
      "epoch": 12.15,
      "learning_rate": 4.296617647058824e-05,
      "loss": 2.9429,
      "step": 226000
    },
    {
      "epoch": 12.2,
      "learning_rate": 4.2672352941176475e-05,
      "loss": 2.9443,
      "step": 227000
    },
    {
      "epoch": 12.26,
      "learning_rate": 4.2378235294117645e-05,
      "loss": 2.9458,
      "step": 228000
    },
    {
      "epoch": 12.31,
      "learning_rate": 4.208411764705882e-05,
      "loss": 2.9486,
      "step": 229000
    },
    {
      "epoch": 12.37,
      "learning_rate": 4.179e-05,
      "loss": 2.9483,
      "step": 230000
    },
    {
      "epoch": 12.42,
      "learning_rate": 4.149617647058824e-05,
      "loss": 2.9469,
      "step": 231000
    },
    {
      "epoch": 12.47,
      "learning_rate": 4.120205882352942e-05,
      "loss": 2.9515,
      "step": 232000
    },
    {
      "epoch": 12.53,
      "learning_rate": 4.090794117647059e-05,
      "loss": 2.9515,
      "step": 233000
    },
    {
      "epoch": 12.58,
      "learning_rate": 4.0614117647058824e-05,
      "loss": 2.9533,
      "step": 234000
    },
    {
      "epoch": 12.63,
      "learning_rate": 4.032e-05,
      "loss": 2.9555,
      "step": 235000
    },
    {
      "epoch": 12.69,
      "learning_rate": 4.002588235294118e-05,
      "loss": 2.9613,
      "step": 236000
    },
    {
      "epoch": 12.74,
      "learning_rate": 3.973205882352941e-05,
      "loss": 2.961,
      "step": 237000
    },
    {
      "epoch": 12.8,
      "learning_rate": 3.943794117647059e-05,
      "loss": 2.9599,
      "step": 238000
    },
    {
      "epoch": 12.85,
      "learning_rate": 3.9143823529411766e-05,
      "loss": 2.964,
      "step": 239000
    },
    {
      "epoch": 12.9,
      "learning_rate": 3.884970588235294e-05,
      "loss": 2.9616,
      "step": 240000
    },
    {
      "epoch": 12.96,
      "learning_rate": 3.855588235294118e-05,
      "loss": 2.9614,
      "step": 241000
    },
    {
      "epoch": 13.0,
      "eval_accuracy": 0.4065392061740851,
      "eval_loss": 3.3943071365356445,
      "eval_runtime": 154.28,
      "eval_samples_per_second": 375.408,
      "eval_steps_per_second": 5.866,
      "step": 241800
    },
    {
      "epoch": 13.01,
      "learning_rate": 3.8261764705882355e-05,
      "loss": 2.9522,
      "step": 242000
    },
    {
      "epoch": 13.06,
      "learning_rate": 3.796794117647059e-05,
      "loss": 2.9162,
      "step": 243000
    },
    {
      "epoch": 13.12,
      "learning_rate": 3.767411764705883e-05,
      "loss": 2.9204,
      "step": 244000
    },
    {
      "epoch": 13.17,
      "learning_rate": 3.7380000000000005e-05,
      "loss": 2.9244,
      "step": 245000
    },
    {
      "epoch": 13.23,
      "learning_rate": 3.7085882352941174e-05,
      "loss": 2.9251,
      "step": 246000
    },
    {
      "epoch": 13.28,
      "learning_rate": 3.679205882352941e-05,
      "loss": 2.9263,
      "step": 247000
    },
    {
      "epoch": 13.33,
      "learning_rate": 3.649794117647059e-05,
      "loss": 2.9264,
      "step": 248000
    },
    {
      "epoch": 13.39,
      "learning_rate": 3.620382352941177e-05,
      "loss": 2.9332,
      "step": 249000
    },
    {
      "epoch": 13.44,
      "learning_rate": 3.590970588235295e-05,
      "loss": 2.9286,
      "step": 250000
    },
    {
      "epoch": 13.49,
      "learning_rate": 3.5615882352941177e-05,
      "loss": 2.9327,
      "step": 251000
    },
    {
      "epoch": 13.55,
      "learning_rate": 3.532176470588235e-05,
      "loss": 2.9358,
      "step": 252000
    },
    {
      "epoch": 13.6,
      "learning_rate": 3.502764705882353e-05,
      "loss": 2.9336,
      "step": 253000
    },
    {
      "epoch": 13.66,
      "learning_rate": 3.4733823529411766e-05,
      "loss": 2.9392,
      "step": 254000
    },
    {
      "epoch": 13.71,
      "learning_rate": 3.443970588235294e-05,
      "loss": 2.9395,
      "step": 255000
    },
    {
      "epoch": 13.76,
      "learning_rate": 3.414558823529412e-05,
      "loss": 2.9401,
      "step": 256000
    },
    {
      "epoch": 13.82,
      "learning_rate": 3.3851764705882355e-05,
      "loss": 2.9357,
      "step": 257000
    },
    {
      "epoch": 13.87,
      "learning_rate": 3.355764705882353e-05,
      "loss": 2.9417,
      "step": 258000
    },
    {
      "epoch": 13.92,
      "learning_rate": 3.326382352941176e-05,
      "loss": 2.9434,
      "step": 259000
    },
    {
      "epoch": 13.98,
      "learning_rate": 3.2969705882352945e-05,
      "loss": 2.9382,
      "step": 260000
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.40722083536055204,
      "eval_loss": 3.389923095703125,
      "eval_runtime": 153.371,
      "eval_samples_per_second": 377.633,
      "eval_steps_per_second": 5.901,
      "step": 260400
    },
    {
      "epoch": 14.03,
      "learning_rate": 3.267558823529412e-05,
      "loss": 2.9117,
      "step": 261000
    },
    {
      "epoch": 14.09,
      "learning_rate": 3.23814705882353e-05,
      "loss": 2.8994,
      "step": 262000
    },
    {
      "epoch": 14.14,
      "learning_rate": 3.2087647058823534e-05,
      "loss": 2.9038,
      "step": 263000
    },
    {
      "epoch": 14.19,
      "learning_rate": 3.1793529411764704e-05,
      "loss": 2.905,
      "step": 264000
    },
    {
      "epoch": 14.25,
      "learning_rate": 3.149970588235294e-05,
      "loss": 2.906,
      "step": 265000
    },
    {
      "epoch": 14.3,
      "learning_rate": 3.1205588235294117e-05,
      "loss": 2.9087,
      "step": 266000
    },
    {
      "epoch": 14.35,
      "learning_rate": 3.09114705882353e-05,
      "loss": 2.9106,
      "step": 267000
    },
    {
      "epoch": 14.41,
      "learning_rate": 3.0617352941176476e-05,
      "loss": 2.9138,
      "step": 268000
    },
    {
      "epoch": 14.46,
      "learning_rate": 3.0323235294117646e-05,
      "loss": 2.9097,
      "step": 269000
    },
    {
      "epoch": 14.52,
      "learning_rate": 3.0029411764705882e-05,
      "loss": 2.9133,
      "step": 270000
    },
    {
      "epoch": 14.57,
      "learning_rate": 2.9735294117647062e-05,
      "loss": 2.9163,
      "step": 271000
    },
    {
      "epoch": 14.62,
      "learning_rate": 2.944117647058824e-05,
      "loss": 2.92,
      "step": 272000
    },
    {
      "epoch": 14.68,
      "learning_rate": 2.914735294117647e-05,
      "loss": 2.919,
      "step": 273000
    },
    {
      "epoch": 14.73,
      "learning_rate": 2.8853235294117648e-05,
      "loss": 2.923,
      "step": 274000
    },
    {
      "epoch": 14.78,
      "learning_rate": 2.8559117647058825e-05,
      "loss": 2.9186,
      "step": 275000
    },
    {
      "epoch": 14.84,
      "learning_rate": 2.8265e-05,
      "loss": 2.9234,
      "step": 276000
    },
    {
      "epoch": 14.89,
      "learning_rate": 2.7971176470588234e-05,
      "loss": 2.9229,
      "step": 277000
    },
    {
      "epoch": 14.95,
      "learning_rate": 2.767735294117647e-05,
      "loss": 2.925,
      "step": 278000
    },
    {
      "epoch": 15.0,
      "learning_rate": 2.7383235294117647e-05,
      "loss": 2.9179,
      "step": 279000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.40748156893759496,
      "eval_loss": 3.392575979232788,
      "eval_runtime": 153.4298,
      "eval_samples_per_second": 377.489,
      "eval_steps_per_second": 5.898,
      "step": 279000
    },
    {
      "epoch": 15.05,
      "learning_rate": 2.7089117647058827e-05,
      "loss": 2.8818,
      "step": 280000
    },
    {
      "epoch": 15.11,
      "learning_rate": 2.6795000000000003e-05,
      "loss": 2.8828,
      "step": 281000
    },
    {
      "epoch": 15.16,
      "learning_rate": 2.6501176470588236e-05,
      "loss": 2.887,
      "step": 282000
    },
    {
      "epoch": 15.22,
      "learning_rate": 2.6207352941176473e-05,
      "loss": 2.8948,
      "step": 283000
    },
    {
      "epoch": 15.27,
      "learning_rate": 2.591323529411765e-05,
      "loss": 2.8877,
      "step": 284000
    },
    {
      "epoch": 15.32,
      "learning_rate": 2.5619117647058826e-05,
      "loss": 2.8937,
      "step": 285000
    },
    {
      "epoch": 15.38,
      "learning_rate": 2.532529411764706e-05,
      "loss": 2.8928,
      "step": 286000
    },
    {
      "epoch": 15.43,
      "learning_rate": 2.5031176470588235e-05,
      "loss": 2.9,
      "step": 287000
    },
    {
      "epoch": 15.48,
      "learning_rate": 2.473735294117647e-05,
      "loss": 2.898,
      "step": 288000
    },
    {
      "epoch": 15.54,
      "learning_rate": 2.4443235294117648e-05,
      "loss": 2.898,
      "step": 289000
    },
    {
      "epoch": 15.59,
      "learning_rate": 2.4149411764705885e-05,
      "loss": 2.8988,
      "step": 290000
    },
    {
      "epoch": 15.65,
      "learning_rate": 2.3855294117647058e-05,
      "loss": 2.9009,
      "step": 291000
    },
    {
      "epoch": 15.7,
      "learning_rate": 2.3561176470588238e-05,
      "loss": 2.9025,
      "step": 292000
    },
    {
      "epoch": 15.75,
      "learning_rate": 2.3267058823529414e-05,
      "loss": 2.8984,
      "step": 293000
    },
    {
      "epoch": 15.81,
      "learning_rate": 2.2972941176470587e-05,
      "loss": 2.899,
      "step": 294000
    },
    {
      "epoch": 15.86,
      "learning_rate": 2.2678823529411767e-05,
      "loss": 2.9063,
      "step": 295000
    },
    {
      "epoch": 15.91,
      "learning_rate": 2.2385e-05,
      "loss": 2.899,
      "step": 296000
    },
    {
      "epoch": 15.97,
      "learning_rate": 2.2090882352941176e-05,
      "loss": 2.9009,
      "step": 297000
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.4071824072219103,
      "eval_loss": 3.404289722442627,
      "eval_runtime": 153.5273,
      "eval_samples_per_second": 377.249,
      "eval_steps_per_second": 5.895,
      "step": 297600
    },
    {
      "epoch": 16.02,
      "learning_rate": 2.1797058823529413e-05,
      "loss": 2.8899,
      "step": 298000
    },
    {
      "epoch": 16.08,
      "learning_rate": 2.150294117647059e-05,
      "loss": 2.8646,
      "step": 299000
    },
    {
      "epoch": 16.13,
      "learning_rate": 2.1209117647058822e-05,
      "loss": 2.871,
      "step": 300000
    },
    {
      "epoch": 16.18,
      "learning_rate": 2.0915000000000002e-05,
      "loss": 2.8711,
      "step": 301000
    },
    {
      "epoch": 16.24,
      "learning_rate": 2.062088235294118e-05,
      "loss": 2.8784,
      "step": 302000
    },
    {
      "epoch": 16.29,
      "learning_rate": 2.032705882352941e-05,
      "loss": 2.8771,
      "step": 303000
    },
    {
      "epoch": 16.34,
      "learning_rate": 2.003294117647059e-05,
      "loss": 2.8738,
      "step": 304000
    },
    {
      "epoch": 16.4,
      "learning_rate": 1.9738823529411765e-05,
      "loss": 2.8755,
      "step": 305000
    },
    {
      "epoch": 16.45,
      "learning_rate": 1.944470588235294e-05,
      "loss": 2.882,
      "step": 306000
    },
    {
      "epoch": 16.51,
      "learning_rate": 1.9150882352941178e-05,
      "loss": 2.8828,
      "step": 307000
    },
    {
      "epoch": 16.56,
      "learning_rate": 1.8856764705882354e-05,
      "loss": 2.8842,
      "step": 308000
    },
    {
      "epoch": 16.61,
      "learning_rate": 1.8562941176470587e-05,
      "loss": 2.8784,
      "step": 309000
    },
    {
      "epoch": 16.67,
      "learning_rate": 1.8268823529411767e-05,
      "loss": 2.8856,
      "step": 310000
    },
    {
      "epoch": 16.72,
      "learning_rate": 1.7975e-05,
      "loss": 2.8857,
      "step": 311000
    },
    {
      "epoch": 16.77,
      "learning_rate": 1.7680882352941176e-05,
      "loss": 2.8827,
      "step": 312000
    },
    {
      "epoch": 16.83,
      "learning_rate": 1.7386764705882353e-05,
      "loss": 2.8874,
      "step": 313000
    },
    {
      "epoch": 16.88,
      "learning_rate": 1.709264705882353e-05,
      "loss": 2.8817,
      "step": 314000
    },
    {
      "epoch": 16.94,
      "learning_rate": 1.6798529411764706e-05,
      "loss": 2.8853,
      "step": 315000
    },
    {
      "epoch": 16.99,
      "learning_rate": 1.6504411764705882e-05,
      "loss": 2.8878,
      "step": 316000
    },
    {
      "epoch": 17.0,
      "eval_accuracy": 0.4078782790052164,
      "eval_loss": 3.395479202270508,
      "eval_runtime": 153.8849,
      "eval_samples_per_second": 376.372,
      "eval_steps_per_second": 5.881,
      "step": 316200
    },
    {
      "epoch": 17.04,
      "learning_rate": 1.621088235294118e-05,
      "loss": 2.8577,
      "step": 317000
    },
    {
      "epoch": 17.1,
      "learning_rate": 1.591705882352941e-05,
      "loss": 2.8562,
      "step": 318000
    },
    {
      "epoch": 17.15,
      "learning_rate": 1.562294117647059e-05,
      "loss": 2.8575,
      "step": 319000
    },
    {
      "epoch": 17.2,
      "learning_rate": 1.5328823529411765e-05,
      "loss": 2.8604,
      "step": 320000
    },
    {
      "epoch": 17.26,
      "learning_rate": 1.5034705882352943e-05,
      "loss": 2.8587,
      "step": 321000
    },
    {
      "epoch": 17.31,
      "learning_rate": 1.4740588235294118e-05,
      "loss": 2.8609,
      "step": 322000
    },
    {
      "epoch": 17.37,
      "learning_rate": 1.4446470588235294e-05,
      "loss": 2.859,
      "step": 323000
    },
    {
      "epoch": 17.42,
      "learning_rate": 1.4152647058823529e-05,
      "loss": 2.8649,
      "step": 324000
    },
    {
      "epoch": 17.47,
      "learning_rate": 1.3858529411764707e-05,
      "loss": 2.8676,
      "step": 325000
    },
    {
      "epoch": 17.53,
      "learning_rate": 1.356470588235294e-05,
      "loss": 2.8672,
      "step": 326000
    },
    {
      "epoch": 17.58,
      "learning_rate": 1.3270588235294118e-05,
      "loss": 2.8657,
      "step": 327000
    },
    {
      "epoch": 17.63,
      "learning_rate": 1.2976764705882354e-05,
      "loss": 2.8713,
      "step": 328000
    },
    {
      "epoch": 17.69,
      "learning_rate": 1.268264705882353e-05,
      "loss": 2.8672,
      "step": 329000
    },
    {
      "epoch": 17.74,
      "learning_rate": 1.2388823529411766e-05,
      "loss": 2.8662,
      "step": 330000
    },
    {
      "epoch": 17.8,
      "learning_rate": 1.2095e-05,
      "loss": 2.8662,
      "step": 331000
    },
    {
      "epoch": 17.85,
      "learning_rate": 1.1800882352941177e-05,
      "loss": 2.8709,
      "step": 332000
    },
    {
      "epoch": 17.9,
      "learning_rate": 1.1507058823529413e-05,
      "loss": 2.8668,
      "step": 333000
    },
    {
      "epoch": 17.96,
      "learning_rate": 1.1212941176470588e-05,
      "loss": 2.8705,
      "step": 334000
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.407813851409137,
      "eval_loss": 3.4078526496887207,
      "eval_runtime": 153.4401,
      "eval_samples_per_second": 377.463,
      "eval_steps_per_second": 5.898,
      "step": 334800
    },
    {
      "epoch": 18.01,
      "learning_rate": 1.0918823529411764e-05,
      "loss": 2.8644,
      "step": 335000
    },
    {
      "epoch": 18.06,
      "learning_rate": 1.0624705882352943e-05,
      "loss": 2.8429,
      "step": 336000
    },
    {
      "epoch": 18.12,
      "learning_rate": 1.0330882352941176e-05,
      "loss": 2.8407,
      "step": 337000
    },
    {
      "epoch": 18.17,
      "learning_rate": 1.0036764705882354e-05,
      "loss": 2.8439,
      "step": 338000
    },
    {
      "epoch": 18.23,
      "learning_rate": 9.742941176470588e-06,
      "loss": 2.8476,
      "step": 339000
    },
    {
      "epoch": 18.28,
      "learning_rate": 9.448823529411765e-06,
      "loss": 2.8499,
      "step": 340000
    },
    {
      "epoch": 18.33,
      "learning_rate": 9.154705882352941e-06,
      "loss": 2.8482,
      "step": 341000
    },
    {
      "epoch": 18.39,
      "learning_rate": 8.860588235294118e-06,
      "loss": 2.8504,
      "step": 342000
    },
    {
      "epoch": 18.44,
      "learning_rate": 8.566764705882353e-06,
      "loss": 2.849,
      "step": 343000
    },
    {
      "epoch": 18.49,
      "learning_rate": 8.272647058823529e-06,
      "loss": 2.8568,
      "step": 344000
    },
    {
      "epoch": 18.55,
      "learning_rate": 7.978823529411765e-06,
      "loss": 2.8513,
      "step": 345000
    },
    {
      "epoch": 18.6,
      "learning_rate": 7.68470588235294e-06,
      "loss": 2.8523,
      "step": 346000
    },
    {
      "epoch": 18.66,
      "learning_rate": 7.3905882352941184e-06,
      "loss": 2.8524,
      "step": 347000
    },
    {
      "epoch": 18.71,
      "learning_rate": 7.096764705882353e-06,
      "loss": 2.8527,
      "step": 348000
    },
    {
      "epoch": 18.76,
      "learning_rate": 6.8026470588235304e-06,
      "loss": 2.8531,
      "step": 349000
    },
    {
      "epoch": 18.82,
      "learning_rate": 6.508529411764706e-06,
      "loss": 2.85,
      "step": 350000
    },
    {
      "epoch": 18.87,
      "learning_rate": 6.214705882352942e-06,
      "loss": 2.8567,
      "step": 351000
    },
    {
      "epoch": 18.92,
      "learning_rate": 5.920588235294118e-06,
      "loss": 2.8546,
      "step": 352000
    },
    {
      "epoch": 18.98,
      "learning_rate": 5.626470588235294e-06,
      "loss": 2.8533,
      "step": 353000
    },
    {
      "epoch": 19.0,
      "eval_accuracy": 0.40773430784943104,
      "eval_loss": 3.4118826389312744,
      "eval_runtime": 153.7844,
      "eval_samples_per_second": 376.618,
      "eval_steps_per_second": 5.885,
      "step": 353400
    },
    {
      "epoch": 19.03,
      "learning_rate": 5.332941176470588e-06,
      "loss": 2.8425,
      "step": 354000
    },
    {
      "epoch": 19.09,
      "learning_rate": 5.038823529411766e-06,
      "loss": 2.8376,
      "step": 355000
    },
    {
      "epoch": 19.14,
      "learning_rate": 4.744705882352941e-06,
      "loss": 2.8358,
      "step": 356000
    },
    {
      "epoch": 19.19,
      "learning_rate": 4.450588235294118e-06,
      "loss": 2.8429,
      "step": 357000
    },
    {
      "epoch": 19.25,
      "learning_rate": 4.156470588235294e-06,
      "loss": 2.8368,
      "step": 358000
    },
    {
      "epoch": 19.3,
      "learning_rate": 3.86264705882353e-06,
      "loss": 2.837,
      "step": 359000
    },
    {
      "epoch": 19.35,
      "learning_rate": 3.568823529411765e-06,
      "loss": 2.8335,
      "step": 360000
    },
    {
      "epoch": 19.41,
      "learning_rate": 3.2747058823529414e-06,
      "loss": 2.8393,
      "step": 361000
    },
    {
      "epoch": 19.46,
      "learning_rate": 2.9805882352941175e-06,
      "loss": 2.8361,
      "step": 362000
    },
    {
      "epoch": 19.52,
      "learning_rate": 2.686470588235294e-06,
      "loss": 2.8365,
      "step": 363000
    },
    {
      "epoch": 19.57,
      "learning_rate": 2.3923529411764705e-06,
      "loss": 2.8391,
      "step": 364000
    },
    {
      "epoch": 19.62,
      "learning_rate": 2.098235294117647e-06,
      "loss": 2.8411,
      "step": 365000
    },
    {
      "epoch": 19.68,
      "learning_rate": 1.8044117647058827e-06,
      "loss": 2.8378,
      "step": 366000
    },
    {
      "epoch": 19.73,
      "learning_rate": 1.510294117647059e-06,
      "loss": 2.837,
      "step": 367000
    },
    {
      "epoch": 19.78,
      "learning_rate": 1.216470588235294e-06,
      "loss": 2.8412,
      "step": 368000
    },
    {
      "epoch": 19.84,
      "learning_rate": 9.223529411764706e-07,
      "loss": 2.8351,
      "step": 369000
    },
    {
      "epoch": 19.89,
      "learning_rate": 6.28235294117647e-07,
      "loss": 2.8366,
      "step": 370000
    },
    {
      "epoch": 19.95,
      "learning_rate": 3.3411764705882354e-07,
      "loss": 2.84,
      "step": 371000
    },
    {
      "epoch": 20.0,
      "learning_rate": 4.029411764705882e-08,
      "loss": 2.8352,
      "step": 372000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.4080045140970133,
      "eval_loss": 3.4137637615203857,
      "eval_runtime": 153.6653,
      "eval_samples_per_second": 376.91,
      "eval_steps_per_second": 5.889,
      "step": 372000
    },
    {
      "epoch": 20.0,
      "step": 372000,
      "total_flos": 1.56732865394688e+18,
      "train_loss": 3.1507886162214382,
      "train_runtime": 81189.7334,
      "train_samples_per_second": 146.617,
      "train_steps_per_second": 4.582
    }
  ],
  "logging_steps": 1000,
  "max_steps": 372000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 5000,
  "total_flos": 1.56732865394688e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}