{
  "best_metric": 0.64125,
  "best_model_checkpoint": "./results/checkpoint-5000",
  "epoch": 25.0,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "learning_rate": 1.9333333333333333e-05,
      "loss": 1.0006,
      "step": 200
    },
    {
      "epoch": 1.25,
      "eval_accuracy": 0.58125,
      "eval_loss": 0.8775196075439453,
      "eval_runtime": 16.4637,
      "eval_samples_per_second": 48.592,
      "eval_steps_per_second": 3.037,
      "step": 250
    },
    {
      "epoch": 2.0,
      "learning_rate": 1.866666666666667e-05,
      "loss": 0.8654,
      "step": 400
    },
    {
      "epoch": 2.5,
      "eval_accuracy": 0.6275,
      "eval_loss": 0.8210939168930054,
      "eval_runtime": 15.3938,
      "eval_samples_per_second": 51.969,
      "eval_steps_per_second": 3.248,
      "step": 500
    },
    {
      "epoch": 3.0,
      "learning_rate": 1.8e-05,
      "loss": 0.7041,
      "step": 600
    },
    {
      "epoch": 3.75,
      "eval_accuracy": 0.62125,
      "eval_loss": 0.9076805710792542,
      "eval_runtime": 15.4386,
      "eval_samples_per_second": 51.818,
      "eval_steps_per_second": 3.239,
      "step": 750
    },
    {
      "epoch": 4.0,
      "learning_rate": 1.7333333333333336e-05,
      "loss": 0.5119,
      "step": 800
    },
    {
      "epoch": 5.0,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.3314,
      "step": 1000
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.63125,
      "eval_loss": 1.0315085649490356,
      "eval_runtime": 15.309,
      "eval_samples_per_second": 52.257,
      "eval_steps_per_second": 3.266,
      "step": 1000
    },
    {
      "epoch": 6.0,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 0.2029,
      "step": 1200
    },
    {
      "epoch": 6.25,
      "eval_accuracy": 0.64,
      "eval_loss": 1.349707841873169,
      "eval_runtime": 15.3657,
      "eval_samples_per_second": 52.064,
      "eval_steps_per_second": 3.254,
      "step": 1250
    },
    {
      "epoch": 7.0,
      "learning_rate": 1.5333333333333334e-05,
      "loss": 0.1332,
      "step": 1400
    },
    {
      "epoch": 7.5,
      "eval_accuracy": 0.56375,
      "eval_loss": 2.2334282398223877,
      "eval_runtime": 15.4014,
      "eval_samples_per_second": 51.943,
      "eval_steps_per_second": 3.246,
      "step": 1500
    },
    {
      "epoch": 8.0,
      "learning_rate": 1.4666666666666666e-05,
      "loss": 0.0748,
      "step": 1600
    },
    {
      "epoch": 8.75,
      "eval_accuracy": 0.605,
      "eval_loss": 2.0466909408569336,
      "eval_runtime": 15.38,
      "eval_samples_per_second": 52.016,
      "eval_steps_per_second": 3.251,
      "step": 1750
    },
    {
      "epoch": 9.0,
      "learning_rate": 1.4e-05,
      "loss": 0.0476,
      "step": 1800
    },
    {
      "epoch": 10.0,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0282,
      "step": 2000
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.6175,
      "eval_loss": 2.3013718128204346,
      "eval_runtime": 15.3412,
      "eval_samples_per_second": 52.147,
      "eval_steps_per_second": 3.259,
      "step": 2000
    },
    {
      "epoch": 11.0,
      "learning_rate": 1.2666666666666667e-05,
      "loss": 0.038,
      "step": 2200
    },
    {
      "epoch": 11.25,
      "eval_accuracy": 0.63,
      "eval_loss": 2.488283157348633,
      "eval_runtime": 15.3496,
      "eval_samples_per_second": 52.119,
      "eval_steps_per_second": 3.257,
      "step": 2250
    },
    {
      "epoch": 12.0,
      "learning_rate": 1.2e-05,
      "loss": 0.0218,
      "step": 2400
    },
    {
      "epoch": 12.5,
      "eval_accuracy": 0.62875,
      "eval_loss": 2.5122828483581543,
      "eval_runtime": 15.4178,
      "eval_samples_per_second": 51.888,
      "eval_steps_per_second": 3.243,
      "step": 2500
    },
    {
      "epoch": 13.0,
      "learning_rate": 1.1333333333333334e-05,
      "loss": 0.0174,
      "step": 2600
    },
    {
      "epoch": 13.75,
      "eval_accuracy": 0.62125,
      "eval_loss": 2.5520098209381104,
      "eval_runtime": 15.4539,
      "eval_samples_per_second": 51.767,
      "eval_steps_per_second": 3.235,
      "step": 2750
    },
    {
      "epoch": 14.0,
      "learning_rate": 1.0666666666666667e-05,
      "loss": 0.0114,
      "step": 2800
    },
    {
      "epoch": 15.0,
      "learning_rate": 1e-05,
      "loss": 0.0134,
      "step": 3000
    },
    {
      "epoch": 15.0,
      "eval_accuracy": 0.61,
      "eval_loss": 2.702610969543457,
      "eval_runtime": 15.3837,
      "eval_samples_per_second": 52.003,
      "eval_steps_per_second": 3.25,
      "step": 3000
    },
    {
      "epoch": 16.0,
      "learning_rate": 9.333333333333334e-06,
      "loss": 0.0107,
      "step": 3200
    },
    {
      "epoch": 16.25,
      "eval_accuracy": 0.605,
      "eval_loss": 2.7754039764404297,
      "eval_runtime": 15.3762,
      "eval_samples_per_second": 52.029,
      "eval_steps_per_second": 3.252,
      "step": 3250
    },
    {
      "epoch": 17.0,
      "learning_rate": 8.666666666666668e-06,
      "loss": 0.0088,
      "step": 3400
    },
    {
      "epoch": 17.5,
      "eval_accuracy": 0.59875,
      "eval_loss": 3.0348358154296875,
      "eval_runtime": 15.3746,
      "eval_samples_per_second": 52.034,
      "eval_steps_per_second": 3.252,
      "step": 3500
    },
    {
      "epoch": 18.0,
      "learning_rate": 8.000000000000001e-06,
      "loss": 0.0084,
      "step": 3600
    },
    {
      "epoch": 18.75,
      "eval_accuracy": 0.6375,
      "eval_loss": 2.828749656677246,
      "eval_runtime": 15.4159,
      "eval_samples_per_second": 51.894,
      "eval_steps_per_second": 3.243,
      "step": 3750
    },
    {
      "epoch": 19.0,
      "learning_rate": 7.333333333333333e-06,
      "loss": 0.0085,
      "step": 3800
    },
    {
      "epoch": 20.0,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.0044,
      "step": 4000
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.62875,
      "eval_loss": 2.9390206336975098,
      "eval_runtime": 15.4137,
      "eval_samples_per_second": 51.902,
      "eval_steps_per_second": 3.244,
      "step": 4000
    },
    {
      "epoch": 21.0,
      "learning_rate": 6e-06,
      "loss": 0.0093,
      "step": 4200
    },
    {
      "epoch": 21.25,
      "eval_accuracy": 0.60375,
      "eval_loss": 3.133685350418091,
      "eval_runtime": 15.4418,
      "eval_samples_per_second": 51.808,
      "eval_steps_per_second": 3.238,
      "step": 4250
    },
    {
      "epoch": 22.0,
      "learning_rate": 5.333333333333334e-06,
      "loss": 0.0071,
      "step": 4400
    },
    {
      "epoch": 22.5,
      "eval_accuracy": 0.6175,
      "eval_loss": 2.979722261428833,
      "eval_runtime": 15.4954,
      "eval_samples_per_second": 51.628,
      "eval_steps_per_second": 3.227,
      "step": 4500
    },
    {
      "epoch": 23.0,
      "learning_rate": 4.666666666666667e-06,
      "loss": 0.0087,
      "step": 4600
    },
    {
      "epoch": 23.75,
      "eval_accuracy": 0.61625,
      "eval_loss": 3.0937910079956055,
      "eval_runtime": 15.473,
      "eval_samples_per_second": 51.703,
      "eval_steps_per_second": 3.231,
      "step": 4750
    },
    {
      "epoch": 24.0,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.0048,
      "step": 4800
    },
    {
      "epoch": 25.0,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.0033,
      "step": 5000
    },
    {
      "epoch": 25.0,
      "eval_accuracy": 0.64125,
      "eval_loss": 2.959096908569336,
      "eval_runtime": 15.3539,
      "eval_samples_per_second": 52.104,
      "eval_steps_per_second": 3.257,
      "step": 5000
    }
  ],
  "max_steps": 6000,
  "num_train_epochs": 30,
  "total_flos": 1.059758088192e+16,
  "trial_name": null,
  "trial_params": null
}