checkpoint-10000/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.195510803440319,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.05,
      "learning_rate": 6.234097519017773e-05,
      "loss": 2.9092,
      "step": 2500
    },
    {
      "epoch": 1.05,
      "eval_loss": 2.7911548614501953,
      "eval_runtime": 359.1547,
      "eval_samples_per_second": 11.8,
      "eval_steps_per_second": 2.951,
      "step": 2500
    },
    {
      "epoch": 2.1,
      "learning_rate": 6.137463890549805e-05,
      "loss": 2.7715,
      "step": 5000
    },
    {
      "epoch": 2.1,
      "eval_loss": 2.7525341510772705,
      "eval_runtime": 358.9958,
      "eval_samples_per_second": 11.805,
      "eval_steps_per_second": 2.953,
      "step": 5000
    },
    {
      "epoch": 3.15,
      "learning_rate": 5.9557478454587004e-05,
      "loss": 2.688,
      "step": 7500
    },
    {
      "epoch": 3.15,
      "eval_loss": 2.737689256668091,
      "eval_runtime": 359.124,
      "eval_samples_per_second": 11.801,
      "eval_steps_per_second": 2.952,
      "step": 7500
    },
    {
      "epoch": 4.2,
      "learning_rate": 5.6939618780421336e-05,
      "loss": 2.6213,
      "step": 10000
    },
    {
      "epoch": 4.2,
      "eval_loss": 2.728829860687256,
      "eval_runtime": 359.1054,
      "eval_samples_per_second": 11.802,
      "eval_steps_per_second": 2.952,
      "step": 10000
    }
  ],
  "max_steps": 47660,
  "num_train_epochs": 20,
  "total_flos": 2.433430888256635e+17,
  "trial_name": null,
  "trial_params": null
}