checkpoint-200/trainer_state.json

{
  "best_metric": 0.6357369422912598,
  "best_model_checkpoint": "experiments/checkpoint-200",
  "epoch": 15.058823529411764,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.75,
      "learning_rate": 2.9999999999999997e-05,
      "loss": 3.2279,
      "step": 10
    },
    {
      "epoch": 1.51,
      "learning_rate": 5.9999999999999995e-05,
      "loss": 3.0585,
      "step": 20
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.699999999999999e-05,
      "loss": 2.6599,
      "step": 30
    },
    {
      "epoch": 3.01,
      "learning_rate": 0.00011399999999999999,
      "loss": 1.9787,
      "step": 40
    },
    {
      "epoch": 3.76,
      "learning_rate": 0.00014099999999999998,
      "loss": 1.378,
      "step": 50
    },
    {
      "epoch": 3.76,
      "eval_loss": 1.1715341806411743,
      "eval_runtime": 12.5908,
      "eval_samples_per_second": 15.885,
      "eval_steps_per_second": 1.986,
      "step": 50
    },
    {
      "epoch": 4.52,
      "learning_rate": 0.00017099999999999998,
      "loss": 1.1071,
      "step": 60
    },
    {
      "epoch": 5.27,
      "learning_rate": 0.000201,
      "loss": 0.9966,
      "step": 70
    },
    {
      "epoch": 6.02,
      "learning_rate": 0.00023099999999999998,
      "loss": 0.9398,
      "step": 80
    },
    {
      "epoch": 6.78,
      "learning_rate": 0.000261,
      "loss": 0.8755,
      "step": 90
    },
    {
      "epoch": 7.53,
      "learning_rate": 0.00029099999999999997,
      "loss": 0.8274,
      "step": 100
    },
    {
      "epoch": 7.53,
      "eval_loss": 0.7494703531265259,
      "eval_runtime": 12.5869,
      "eval_samples_per_second": 15.89,
      "eval_steps_per_second": 1.986,
      "step": 100
    },
    {
      "epoch": 8.28,
      "learning_rate": 0.0002895,
      "loss": 0.761,
      "step": 110
    },
    {
      "epoch": 9.04,
      "learning_rate": 0.0002745,
      "loss": 0.7358,
      "step": 120
    },
    {
      "epoch": 9.79,
      "learning_rate": 0.00025949999999999997,
      "loss": 0.7155,
      "step": 130
    },
    {
      "epoch": 10.54,
      "learning_rate": 0.0002445,
      "loss": 0.6811,
      "step": 140
    },
    {
      "epoch": 11.29,
      "learning_rate": 0.0002295,
      "loss": 0.6712,
      "step": 150
    },
    {
      "epoch": 11.29,
      "eval_loss": 0.6565232872962952,
      "eval_runtime": 12.5854,
      "eval_samples_per_second": 15.891,
      "eval_steps_per_second": 1.986,
      "step": 150
    },
    {
      "epoch": 12.05,
      "learning_rate": 0.00021449999999999998,
      "loss": 0.6562,
      "step": 160
    },
    {
      "epoch": 12.8,
      "learning_rate": 0.0001995,
      "loss": 0.6367,
      "step": 170
    },
    {
      "epoch": 13.55,
      "learning_rate": 0.00018449999999999999,
      "loss": 0.6173,
      "step": 180
    },
    {
      "epoch": 14.31,
      "learning_rate": 0.00016949999999999997,
      "loss": 0.6267,
      "step": 190
    },
    {
      "epoch": 15.06,
      "learning_rate": 0.0001545,
      "loss": 0.6039,
      "step": 200
    },
    {
      "epoch": 15.06,
      "eval_loss": 0.6357369422912598,
      "eval_runtime": 12.6199,
      "eval_samples_per_second": 15.848,
      "eval_steps_per_second": 1.981,
      "step": 200
    }
  ],
  "max_steps": 300,
  "num_train_epochs": 24,
  "total_flos": 1.3818886480429056e+17,
  "trial_name": null,
  "trial_params": null
}