{
  "best_metric": 0.5172018348623854,
  "best_model_checkpoint": "outputs/soft_prompt/deberta-v2-xlarge/sst2/checkpoint-3400",
  "epoch": 3.0,
  "global_step": 25257,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.831559658050537,
      "eval_runtime": 12.4014,
      "eval_samples_per_second": 70.315,
      "eval_steps_per_second": 8.789,
      "step": 200
    },
    {
      "epoch": 0.05,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7252020835876465,
      "eval_runtime": 12.5411,
      "eval_samples_per_second": 69.532,
      "eval_steps_per_second": 8.691,
      "step": 400
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0294061052381518,
      "loss": 2.7746,
      "step": 500
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 6.318162441253662,
      "eval_runtime": 12.2524,
      "eval_samples_per_second": 71.17,
      "eval_steps_per_second": 8.896,
      "step": 600
    },
    {
      "epoch": 0.1,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.9581830501556396,
      "eval_runtime": 13.2254,
      "eval_samples_per_second": 65.934,
      "eval_steps_per_second": 8.242,
      "step": 800
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.0288122104763036,
      "loss": 2.8668,
      "step": 1000
    },
    {
      "epoch": 0.12,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.97942715883255,
      "eval_runtime": 12.2204,
      "eval_samples_per_second": 71.356,
      "eval_steps_per_second": 8.92,
      "step": 1000
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 9.121758460998535,
      "eval_runtime": 12.6821,
      "eval_samples_per_second": 68.758,
      "eval_steps_per_second": 8.595,
      "step": 1200
    },
    {
      "epoch": 0.17,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.1010525226593018,
      "eval_runtime": 13.3142,
      "eval_samples_per_second": 65.494,
      "eval_steps_per_second": 8.187,
      "step": 1400
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.0282183157144554,
      "loss": 3.0595,
      "step": 1500
    },
    {
      "epoch": 0.19,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.240478754043579,
      "eval_runtime": 11.7539,
      "eval_samples_per_second": 74.188,
      "eval_steps_per_second": 9.274,
      "step": 1600
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.822434663772583,
      "eval_runtime": 11.937,
      "eval_samples_per_second": 73.05,
      "eval_steps_per_second": 9.131,
      "step": 1800
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.027624420952607195,
      "loss": 2.7406,
      "step": 2000
    },
    {
      "epoch": 0.24,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.2581450939178467,
      "eval_runtime": 12.944,
      "eval_samples_per_second": 67.367,
      "eval_steps_per_second": 8.421,
      "step": 2000
    },
    {
      "epoch": 0.26,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 5.0679168701171875,
      "eval_runtime": 12.8993,
      "eval_samples_per_second": 67.6,
      "eval_steps_per_second": 8.45,
      "step": 2200
    },
    {
      "epoch": 0.29,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 4.079117774963379,
      "eval_runtime": 12.6438,
      "eval_samples_per_second": 68.966,
      "eval_steps_per_second": 8.621,
      "step": 2400
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.027030526190758998,
      "loss": 2.341,
      "step": 2500
    },
    {
      "epoch": 0.31,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 6.498974800109863,
      "eval_runtime": 11.7875,
      "eval_samples_per_second": 73.977,
      "eval_steps_per_second": 9.247,
      "step": 2600
    },
    {
      "epoch": 0.33,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.2673330307006836,
      "eval_runtime": 12.4045,
      "eval_samples_per_second": 70.297,
      "eval_steps_per_second": 8.787,
      "step": 2800
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.026436631428910798,
      "loss": 2.5017,
      "step": 3000
    },
    {
      "epoch": 0.36,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.775005340576172,
      "eval_runtime": 12.8935,
      "eval_samples_per_second": 67.631,
      "eval_steps_per_second": 8.454,
      "step": 3000
    },
    {
      "epoch": 0.38,
      "eval_accuracy": 0.5126146788990825,
      "eval_loss": 0.8905919790267944,
      "eval_runtime": 11.7168,
      "eval_samples_per_second": 74.423,
      "eval_steps_per_second": 9.303,
      "step": 3200
    },
    {
      "epoch": 0.4,
      "eval_accuracy": 0.5172018348623854,
      "eval_loss": 3.1630539894104004,
      "eval_runtime": 12.3677,
      "eval_samples_per_second": 70.506,
      "eval_steps_per_second": 8.813,
      "step": 3400
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.025842736667062594,
      "loss": 2.587,
      "step": 3500
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.8831442594528198,
      "eval_runtime": 12.7491,
      "eval_samples_per_second": 68.397,
      "eval_steps_per_second": 8.55,
      "step": 3600
    },
    {
      "epoch": 0.45,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.40650475025177,
      "eval_runtime": 11.8781,
      "eval_samples_per_second": 73.412,
      "eval_steps_per_second": 9.177,
      "step": 3800
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.025248841905214398,
      "loss": 2.2272,
      "step": 4000
    },
    {
      "epoch": 0.48,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 5.7484235763549805,
      "eval_runtime": 12.2084,
      "eval_samples_per_second": 71.426,
      "eval_steps_per_second": 8.928,
      "step": 4000
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 3.051417589187622,
      "eval_runtime": 13.0485,
      "eval_samples_per_second": 66.828,
      "eval_steps_per_second": 8.353,
      "step": 4200
    },
    {
      "epoch": 0.52,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.1735517978668213,
      "eval_runtime": 12.52,
      "eval_samples_per_second": 69.648,
      "eval_steps_per_second": 8.706,
      "step": 4400
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.024654947143366194,
      "loss": 2.474,
      "step": 4500
    },
    {
      "epoch": 0.55,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 3.5813305377960205,
      "eval_runtime": 12.4532,
      "eval_samples_per_second": 70.022,
      "eval_steps_per_second": 8.753,
      "step": 4600
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 3.0905096530914307,
      "eval_runtime": 12.1764,
      "eval_samples_per_second": 71.614,
      "eval_steps_per_second": 8.952,
      "step": 4800
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.024061052381517994,
      "loss": 2.6958,
      "step": 5000
    },
    {
      "epoch": 0.59,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.6613879203796387,
      "eval_runtime": 13.0709,
      "eval_samples_per_second": 66.713,
      "eval_steps_per_second": 8.339,
      "step": 5000
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.9912046194076538,
      "eval_runtime": 11.0684,
      "eval_samples_per_second": 78.783,
      "eval_steps_per_second": 9.848,
      "step": 5200
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.708740711212158,
      "eval_runtime": 12.3098,
      "eval_samples_per_second": 70.838,
      "eval_steps_per_second": 8.855,
      "step": 5400
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.023467157619669794,
      "loss": 2.4571,
      "step": 5500
    },
    {
      "epoch": 0.67,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.9453651309013367,
      "eval_runtime": 12.5807,
      "eval_samples_per_second": 69.313,
      "eval_steps_per_second": 8.664,
      "step": 5600
    },
    {
      "epoch": 0.69,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.9460629820823669,
      "eval_runtime": 12.3033,
      "eval_samples_per_second": 70.875,
      "eval_steps_per_second": 8.859,
      "step": 5800
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.022873262857821593,
      "loss": 2.4046,
      "step": 6000
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.5034403669724771,
      "eval_loss": 0.7613060474395752,
      "eval_runtime": 12.922,
      "eval_samples_per_second": 67.482,
      "eval_steps_per_second": 8.435,
      "step": 6000
    },
    {
      "epoch": 0.74,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 3.473540782928467,
      "eval_runtime": 12.5608,
      "eval_samples_per_second": 69.423,
      "eval_steps_per_second": 8.678,
      "step": 6200
    },
    {
      "epoch": 0.76,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8309389352798462,
      "eval_runtime": 12.3028,
      "eval_samples_per_second": 70.878,
      "eval_steps_per_second": 8.86,
      "step": 6400
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.022279368095973393,
      "loss": 1.9778,
      "step": 6500
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.345905303955078,
      "eval_runtime": 12.6993,
      "eval_samples_per_second": 68.665,
      "eval_steps_per_second": 8.583,
      "step": 6600
    },
    {
      "epoch": 0.81,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.992404818534851,
      "eval_runtime": 11.8953,
      "eval_samples_per_second": 73.306,
      "eval_steps_per_second": 9.163,
      "step": 6800
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.02168547333412519,
      "loss": 1.9132,
      "step": 7000
    },
    {
      "epoch": 0.83,
      "eval_accuracy": 0.48394495412844035,
      "eval_loss": 0.7653124928474426,
      "eval_runtime": 12.5926,
      "eval_samples_per_second": 69.247,
      "eval_steps_per_second": 8.656,
      "step": 7000
    },
    {
      "epoch": 0.86,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.6025058031082153,
      "eval_runtime": 12.8694,
      "eval_samples_per_second": 67.757,
      "eval_steps_per_second": 8.47,
      "step": 7200
    },
    {
      "epoch": 0.88,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.3198583126068115,
      "eval_runtime": 12.2526,
      "eval_samples_per_second": 71.168,
      "eval_steps_per_second": 8.896,
      "step": 7400
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.021091578572276993,
      "loss": 2.1041,
      "step": 7500
    },
    {
      "epoch": 0.9,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.9014425277709961,
      "eval_runtime": 12.1923,
      "eval_samples_per_second": 71.52,
      "eval_steps_per_second": 8.94,
      "step": 7600
    },
    {
      "epoch": 0.93,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.1414848566055298,
      "eval_runtime": 12.2534,
      "eval_samples_per_second": 71.164,
      "eval_steps_per_second": 8.896,
      "step": 7800
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.020497683810428793,
      "loss": 2.2236,
      "step": 8000
    },
    {
      "epoch": 0.95,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.9049626588821411,
      "eval_runtime": 12.5976,
      "eval_samples_per_second": 69.219,
      "eval_steps_per_second": 8.652,
      "step": 8000
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.53812575340271,
      "eval_runtime": 12.192,
      "eval_samples_per_second": 71.522,
      "eval_steps_per_second": 8.94,
      "step": 8200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 3.9090523719787598,
      "eval_runtime": 12.1436,
      "eval_samples_per_second": 71.807,
      "eval_steps_per_second": 8.976,
      "step": 8400
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.01990378904858059,
      "loss": 1.9257,
      "step": 8500
    },
    {
      "epoch": 1.02,
      "eval_accuracy": 0.44380733944954126,
      "eval_loss": 1.3826260566711426,
      "eval_runtime": 12.394,
      "eval_samples_per_second": 70.357,
      "eval_steps_per_second": 8.795,
      "step": 8600
    },
    {
      "epoch": 1.05,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7106401920318604,
      "eval_runtime": 12.254,
      "eval_samples_per_second": 71.161,
      "eval_steps_per_second": 8.895,
      "step": 8800
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.019309894286732392,
      "loss": 1.9533,
      "step": 9000
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.6487476825714111,
      "eval_runtime": 13.5392,
      "eval_samples_per_second": 64.406,
      "eval_steps_per_second": 8.051,
      "step": 9000
    },
    {
      "epoch": 1.09,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.4527719020843506,
      "eval_runtime": 12.2271,
      "eval_samples_per_second": 71.317,
      "eval_steps_per_second": 8.915,
      "step": 9200
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.1699163913726807,
      "eval_runtime": 12.1388,
      "eval_samples_per_second": 71.836,
      "eval_steps_per_second": 8.979,
      "step": 9400
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.01871599952488419,
      "loss": 1.7969,
      "step": 9500
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.155882716178894,
      "eval_runtime": 12.1707,
      "eval_samples_per_second": 71.648,
      "eval_steps_per_second": 8.956,
      "step": 9600
    },
    {
      "epoch": 1.16,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.0576764345169067,
      "eval_runtime": 12.3024,
      "eval_samples_per_second": 70.88,
      "eval_steps_per_second": 8.86,
      "step": 9800
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.01812210476303599,
      "loss": 1.8048,
      "step": 10000
    },
    {
      "epoch": 1.19,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.1585994958877563,
      "eval_runtime": 12.8412,
      "eval_samples_per_second": 67.906,
      "eval_steps_per_second": 8.488,
      "step": 10000
    },
    {
      "epoch": 1.21,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7068227529525757,
      "eval_runtime": 12.4262,
      "eval_samples_per_second": 70.174,
      "eval_steps_per_second": 8.772,
      "step": 10200
    },
    {
      "epoch": 1.24,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7638933658599854,
      "eval_runtime": 12.0016,
      "eval_samples_per_second": 72.657,
      "eval_steps_per_second": 9.082,
      "step": 10400
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.01752821000118779,
      "loss": 2.0729,
      "step": 10500
    },
    {
      "epoch": 1.26,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.188310146331787,
      "eval_runtime": 13.0443,
      "eval_samples_per_second": 66.849,
      "eval_steps_per_second": 8.356,
      "step": 10600
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.6973594427108765,
      "eval_runtime": 11.5707,
      "eval_samples_per_second": 75.363,
      "eval_steps_per_second": 9.42,
      "step": 10800
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.016934315239339588,
      "loss": 1.7558,
      "step": 11000
    },
    {
      "epoch": 1.31,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.8548436164855957,
      "eval_runtime": 13.0206,
      "eval_samples_per_second": 66.971,
      "eval_steps_per_second": 8.371,
      "step": 11000
    },
    {
      "epoch": 1.33,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7331790924072266,
      "eval_runtime": 12.107,
      "eval_samples_per_second": 72.024,
      "eval_steps_per_second": 9.003,
      "step": 11200
    },
    {
      "epoch": 1.35,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.7065454721450806,
      "eval_runtime": 12.0677,
      "eval_samples_per_second": 72.259,
      "eval_steps_per_second": 9.032,
      "step": 11400
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.016340420477491388,
      "loss": 1.931,
      "step": 11500
    },
    {
      "epoch": 1.38,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 3.5782742500305176,
      "eval_runtime": 12.3183,
      "eval_samples_per_second": 70.789,
      "eval_steps_per_second": 8.849,
      "step": 11600
    },
    {
      "epoch": 1.4,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.2677643299102783,
      "eval_runtime": 11.6938,
      "eval_samples_per_second": 74.569,
      "eval_steps_per_second": 9.321,
      "step": 11800
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.015746525715643188,
      "loss": 1.739,
      "step": 12000
    },
    {
      "epoch": 1.43,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.367167353630066,
      "eval_runtime": 12.6396,
      "eval_samples_per_second": 68.99,
      "eval_steps_per_second": 8.624,
      "step": 12000
    },
    {
      "epoch": 1.45,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.4737409353256226,
      "eval_runtime": 12.3657,
      "eval_samples_per_second": 70.517,
      "eval_steps_per_second": 8.815,
      "step": 12200
    },
    {
      "epoch": 1.47,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8329254984855652,
      "eval_runtime": 12.1625,
      "eval_samples_per_second": 71.696,
      "eval_steps_per_second": 8.962,
      "step": 12400
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.015152630953794988,
      "loss": 1.655,
      "step": 12500
    },
    {
      "epoch": 1.5,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.237131118774414,
      "eval_runtime": 12.2197,
      "eval_samples_per_second": 71.36,
      "eval_steps_per_second": 8.92,
      "step": 12600
    },
    {
      "epoch": 1.52,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 3.846944570541382,
      "eval_runtime": 12.0155,
      "eval_samples_per_second": 72.573,
      "eval_steps_per_second": 9.072,
      "step": 12800
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.014558736191946788,
      "loss": 1.7284,
      "step": 13000
    },
    {
      "epoch": 1.54,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.403191328048706,
      "eval_runtime": 12.6645,
      "eval_samples_per_second": 68.854,
      "eval_steps_per_second": 8.607,
      "step": 13000
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.1265949010849,
      "eval_runtime": 12.3046,
      "eval_samples_per_second": 70.868,
      "eval_steps_per_second": 8.859,
      "step": 13200
    },
    {
      "epoch": 1.59,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.9130467176437378,
      "eval_runtime": 12.0511,
      "eval_samples_per_second": 72.359,
      "eval_steps_per_second": 9.045,
      "step": 13400
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.013964841430098586,
      "loss": 1.5742,
      "step": 13500
    },
    {
      "epoch": 1.62,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.832761824131012,
      "eval_runtime": 12.2374,
      "eval_samples_per_second": 71.257,
      "eval_steps_per_second": 8.907,
      "step": 13600
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 3.8501062393188477,
      "eval_runtime": 12.573,
      "eval_samples_per_second": 69.355,
      "eval_steps_per_second": 8.669,
      "step": 13800
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.013370946668250385,
      "loss": 1.7039,
      "step": 14000
    },
    {
      "epoch": 1.66,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.2914493083953857,
      "eval_runtime": 12.0326,
      "eval_samples_per_second": 72.47,
      "eval_steps_per_second": 9.059,
      "step": 14000
    },
    {
      "epoch": 1.69,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.136923909187317,
      "eval_runtime": 11.7256,
      "eval_samples_per_second": 74.367,
      "eval_steps_per_second": 9.296,
      "step": 14200
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7574475407600403,
      "eval_runtime": 12.2584,
      "eval_samples_per_second": 71.135,
      "eval_steps_per_second": 8.892,
      "step": 14400
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.012777051906402184,
      "loss": 1.4352,
      "step": 14500
    },
    {
      "epoch": 1.73,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7623356580734253,
      "eval_runtime": 12.706,
      "eval_samples_per_second": 68.629,
      "eval_steps_per_second": 8.579,
      "step": 14600
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.6579828262329102,
      "eval_runtime": 12.16,
      "eval_samples_per_second": 71.71,
      "eval_steps_per_second": 8.964,
      "step": 14800
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.012183157144553985,
      "loss": 1.6328,
      "step": 15000
    },
    {
      "epoch": 1.78,
      "eval_accuracy": 0.481651376146789,
      "eval_loss": 0.693511426448822,
      "eval_runtime": 12.7779,
      "eval_samples_per_second": 68.243,
      "eval_steps_per_second": 8.53,
      "step": 15000
    },
    {
      "epoch": 1.81,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.6989684700965881,
      "eval_runtime": 12.0393,
      "eval_samples_per_second": 72.43,
      "eval_steps_per_second": 9.054,
      "step": 15200
    },
    {
      "epoch": 1.83,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7303033471107483,
      "eval_runtime": 12.3696,
      "eval_samples_per_second": 70.496,
      "eval_steps_per_second": 8.812,
      "step": 15400
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.011589262382705785,
      "loss": 1.4498,
      "step": 15500
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.675624132156372,
      "eval_runtime": 12.6317,
      "eval_samples_per_second": 69.032,
      "eval_steps_per_second": 8.629,
      "step": 15600
    },
    {
      "epoch": 1.88,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.0083491802215576,
      "eval_runtime": 12.349,
      "eval_samples_per_second": 70.613,
      "eval_steps_per_second": 8.827,
      "step": 15800
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.010995367620857583,
      "loss": 1.4022,
      "step": 16000
    },
    {
      "epoch": 1.9,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7790195345878601,
      "eval_runtime": 11.9659,
      "eval_samples_per_second": 72.874,
      "eval_steps_per_second": 9.109,
      "step": 16000
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.296112060546875,
      "eval_runtime": 13.0796,
      "eval_samples_per_second": 66.669,
      "eval_steps_per_second": 8.334,
      "step": 16200
    },
    {
      "epoch": 1.95,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7208316326141357,
      "eval_runtime": 12.4652,
      "eval_samples_per_second": 69.955,
      "eval_steps_per_second": 8.744,
      "step": 16400
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.010401472859009383,
      "loss": 1.4503,
      "step": 16500
    },
    {
      "epoch": 1.97,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8011333346366882,
      "eval_runtime": 12.77,
      "eval_samples_per_second": 68.285,
      "eval_steps_per_second": 8.536,
      "step": 16600
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8194194436073303,
      "eval_runtime": 12.2506,
      "eval_samples_per_second": 71.18,
      "eval_steps_per_second": 8.898,
      "step": 16800
    },
    {
      "epoch": 2.02,
      "learning_rate": 0.009807578097161183,
      "loss": 1.3401,
      "step": 17000
    },
    {
      "epoch": 2.02,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.9209619760513306,
      "eval_runtime": 12.7245,
      "eval_samples_per_second": 68.529,
      "eval_steps_per_second": 8.566,
      "step": 17000
    },
    {
      "epoch": 2.04,
      "eval_accuracy": 0.4954128440366973,
      "eval_loss": 2.172947645187378,
      "eval_runtime": 13.3907,
      "eval_samples_per_second": 65.12,
      "eval_steps_per_second": 8.14,
      "step": 17200
    },
    {
      "epoch": 2.07,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 2.8217873573303223,
      "eval_runtime": 12.0414,
      "eval_samples_per_second": 72.417,
      "eval_steps_per_second": 9.052,
      "step": 17400
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.009213683335312983,
      "loss": 1.515,
      "step": 17500
    },
    {
      "epoch": 2.09,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.8880350589752197,
      "eval_runtime": 12.7892,
      "eval_samples_per_second": 68.183,
      "eval_steps_per_second": 8.523,
      "step": 17600
    },
    {
      "epoch": 2.11,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8059159517288208,
      "eval_runtime": 12.8563,
      "eval_samples_per_second": 67.827,
      "eval_steps_per_second": 8.478,
      "step": 17800
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.008619788573464782,
      "loss": 1.2519,
      "step": 18000
    },
    {
      "epoch": 2.14,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.1668144464492798,
      "eval_runtime": 12.0845,
      "eval_samples_per_second": 72.158,
      "eval_steps_per_second": 9.02,
      "step": 18000
    },
    {
      "epoch": 2.16,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.8440486192703247,
      "eval_runtime": 11.5453,
      "eval_samples_per_second": 75.529,
      "eval_steps_per_second": 9.441,
      "step": 18200
    },
    {
      "epoch": 2.19,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.482177972793579,
      "eval_runtime": 11.5533,
      "eval_samples_per_second": 75.476,
      "eval_steps_per_second": 9.435,
      "step": 18400
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.008025893811616582,
      "loss": 1.2221,
      "step": 18500
    },
    {
      "epoch": 2.21,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.6977333426475525,
      "eval_runtime": 12.3902,
      "eval_samples_per_second": 70.378,
      "eval_steps_per_second": 8.797,
      "step": 18600
    },
    {
      "epoch": 2.23,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.3418121337890625,
      "eval_runtime": 12.4195,
      "eval_samples_per_second": 70.212,
      "eval_steps_per_second": 8.776,
      "step": 18800
    },
    {
      "epoch": 2.26,
      "learning_rate": 0.00743199904976838,
      "loss": 1.1201,
      "step": 19000
    },
    {
      "epoch": 2.26,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7915144562721252,
      "eval_runtime": 12.717,
      "eval_samples_per_second": 68.57,
      "eval_steps_per_second": 8.571,
      "step": 19000
    },
    {
      "epoch": 2.28,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.961918830871582,
      "eval_runtime": 13.173,
      "eval_samples_per_second": 66.196,
      "eval_steps_per_second": 8.275,
      "step": 19200
    },
    {
      "epoch": 2.3,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.220428228378296,
      "eval_runtime": 12.2958,
      "eval_samples_per_second": 70.918,
      "eval_steps_per_second": 8.865,
      "step": 19400
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.006838104287920181,
      "loss": 1.0869,
      "step": 19500
    },
    {
      "epoch": 2.33,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.6541168689727783,
      "eval_runtime": 12.5954,
      "eval_samples_per_second": 69.232,
      "eval_steps_per_second": 8.654,
      "step": 19600
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 2.7402684688568115,
      "eval_runtime": 12.7516,
      "eval_samples_per_second": 68.383,
      "eval_steps_per_second": 8.548,
      "step": 19800
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.00624420952607198,
      "loss": 1.0804,
      "step": 20000
    },
    {
      "epoch": 2.38,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.2037415504455566,
      "eval_runtime": 13.138,
      "eval_samples_per_second": 66.372,
      "eval_steps_per_second": 8.297,
      "step": 20000
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7337152361869812,
      "eval_runtime": 11.6201,
      "eval_samples_per_second": 75.043,
      "eval_steps_per_second": 9.38,
      "step": 20200
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.8853695392608643,
      "eval_runtime": 13.0127,
      "eval_samples_per_second": 67.012,
      "eval_steps_per_second": 8.376,
      "step": 20400
    },
    {
      "epoch": 2.43,
      "learning_rate": 0.005650314764223779,
      "loss": 1.0025,
      "step": 20500
    },
    {
      "epoch": 2.45,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7113233804702759,
      "eval_runtime": 12.2168,
      "eval_samples_per_second": 71.377,
      "eval_steps_per_second": 8.922,
      "step": 20600
    },
    {
      "epoch": 2.47,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.0583016872406006,
      "eval_runtime": 12.0017,
      "eval_samples_per_second": 72.657,
      "eval_steps_per_second": 9.082,
      "step": 20800
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.005056420002375579,
      "loss": 0.9856,
      "step": 21000
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7113476395606995,
      "eval_runtime": 12.9697,
      "eval_samples_per_second": 67.234,
      "eval_steps_per_second": 8.404,
      "step": 21000
    },
    {
      "epoch": 2.52,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7482680678367615,
      "eval_runtime": 12.7958,
      "eval_samples_per_second": 68.147,
      "eval_steps_per_second": 8.518,
      "step": 21200
    },
    {
      "epoch": 2.54,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.6966097950935364,
      "eval_runtime": 12.2964,
      "eval_samples_per_second": 70.915,
      "eval_steps_per_second": 8.864,
      "step": 21400
    },
    {
      "epoch": 2.55,
      "learning_rate": 0.004462525240527379,
      "loss": 1.0364,
      "step": 21500
    },
    {
      "epoch": 2.57,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.0607110261917114,
      "eval_runtime": 12.067,
      "eval_samples_per_second": 72.263,
      "eval_steps_per_second": 9.033,
      "step": 21600
    },
    {
      "epoch": 2.59,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.1381345987319946,
      "eval_runtime": 12.2151,
      "eval_samples_per_second": 71.387,
      "eval_steps_per_second": 8.923,
      "step": 21800
    },
    {
      "epoch": 2.61,
      "learning_rate": 0.003868630478679178,
      "loss": 0.9683,
      "step": 22000
    },
    {
      "epoch": 2.61,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7138826847076416,
      "eval_runtime": 12.7162,
      "eval_samples_per_second": 68.574,
      "eval_steps_per_second": 8.572,
      "step": 22000
    },
    {
      "epoch": 2.64,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.6361145973205566,
      "eval_runtime": 11.8698,
      "eval_samples_per_second": 73.464,
      "eval_steps_per_second": 9.183,
      "step": 22200
    },
    {
      "epoch": 2.66,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 1.1421782970428467,
      "eval_runtime": 11.321,
      "eval_samples_per_second": 77.025,
      "eval_steps_per_second": 9.628,
      "step": 22400
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.0032747357168309774,
      "loss": 0.908,
      "step": 22500
    },
    {
      "epoch": 2.68,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7317955493927002,
      "eval_runtime": 12.2644,
      "eval_samples_per_second": 71.1,
      "eval_steps_per_second": 8.888,
      "step": 22600
    },
    {
      "epoch": 2.71,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.6962340474128723,
      "eval_runtime": 11.822,
      "eval_samples_per_second": 73.761,
      "eval_steps_per_second": 9.22,
      "step": 22800
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.0026808409549827768,
      "loss": 0.8761,
      "step": 23000
    },
    {
      "epoch": 2.73,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.9568504691123962,
      "eval_runtime": 12.3037,
      "eval_samples_per_second": 70.873,
      "eval_steps_per_second": 8.859,
      "step": 23000
    },
    {
      "epoch": 2.76,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.9744265675544739,
      "eval_runtime": 11.9452,
      "eval_samples_per_second": 73.0,
      "eval_steps_per_second": 9.125,
      "step": 23200
    },
    {
      "epoch": 2.78,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 1.091610312461853,
      "eval_runtime": 12.8562,
      "eval_samples_per_second": 67.827,
      "eval_steps_per_second": 8.478,
      "step": 23400
    },
    {
      "epoch": 2.79,
      "learning_rate": 0.0020869461931345766,
      "loss": 0.8209,
      "step": 23500
    },
    {
      "epoch": 2.8,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.8106526136398315,
      "eval_runtime": 11.7827,
      "eval_samples_per_second": 74.007,
      "eval_steps_per_second": 9.251,
      "step": 23600
    },
    {
      "epoch": 2.83,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7210954427719116,
      "eval_runtime": 11.3303,
      "eval_samples_per_second": 76.962,
      "eval_steps_per_second": 9.62,
      "step": 23800
    },
    {
      "epoch": 2.85,
      "learning_rate": 0.001493051431286376,
      "loss": 0.8008,
      "step": 24000
    },
    {
      "epoch": 2.85,
      "eval_accuracy": 0.49770642201834864,
      "eval_loss": 0.6930689811706543,
      "eval_runtime": 12.4553,
      "eval_samples_per_second": 70.01,
      "eval_steps_per_second": 8.751,
      "step": 24000
    },
    {
      "epoch": 2.87,
      "eval_accuracy": 0.4908256880733945,
      "eval_loss": 0.7184925675392151,
      "eval_runtime": 12.3113,
      "eval_samples_per_second": 70.829,
      "eval_steps_per_second": 8.854,
      "step": 24200
    },
    {
      "epoch": 2.9,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.729166567325592,
      "eval_runtime": 11.2952,
      "eval_samples_per_second": 77.201,
      "eval_steps_per_second": 9.65,
      "step": 24400
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.0008991566694381756,
      "loss": 0.7738,
      "step": 24500
    },
    {
      "epoch": 2.92,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7178325057029724,
      "eval_runtime": 12.4374,
      "eval_samples_per_second": 70.111,
      "eval_steps_per_second": 8.764,
      "step": 24600
    },
    {
      "epoch": 2.95,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.6961101293563843,
      "eval_runtime": 12.5193,
      "eval_samples_per_second": 69.652,
      "eval_steps_per_second": 8.707,
      "step": 24800
    },
    {
      "epoch": 2.97,
      "learning_rate": 0.00030526190758997505,
      "loss": 0.755,
      "step": 25000
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7101095914840698,
      "eval_runtime": 12.5702,
      "eval_samples_per_second": 69.371,
      "eval_steps_per_second": 8.671,
      "step": 25000
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.5091743119266054,
      "eval_loss": 0.7049440145492554,
      "eval_runtime": 11.2857,
      "eval_samples_per_second": 77.266,
      "eval_steps_per_second": 9.658,
      "step": 25200
    },
    {
      "epoch": 3.0,
      "step": 25257,
      "total_flos": 1.070008779996841e+17,
      "train_loss": 1.708119561667774,
      "train_runtime": 7791.0638,
      "train_samples_per_second": 25.933,
      "train_steps_per_second": 3.242
    }
  ],
  "max_steps": 25257,
  "num_train_epochs": 3,
  "total_flos": 1.070008779996841e+17,
  "trial_name": null,
  "trial_params": null
}