trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.3928847641144624,
  "eval_steps": 500,
  "global_step": 12,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "learning_rate": 6.666666666666667e-06,
      "loss": 1.4496,
      "step": 1
    },
    {
      "epoch": 0.4,
      "eval_loss": 1.4424513578414917,
      "eval_runtime": 12.4902,
      "eval_samples_per_second": 46.036,
      "eval_steps_per_second": 5.765,
      "step": 4
    },
    {
      "epoch": 1.1,
      "learning_rate": 1.973044870579824e-05,
      "loss": 1.4508,
      "step": 5
    },
    {
      "epoch": 1.39,
      "eval_loss": 1.437177062034607,
      "eval_runtime": 12.4739,
      "eval_samples_per_second": 46.096,
      "eval_steps_per_second": 5.772,
      "step": 8
    },
    {
      "epoch": 2.19,
      "learning_rate": 1.686241637868734e-05,
      "loss": 1.4458,
      "step": 10
    },
    {
      "epoch": 2.39,
      "eval_loss": 1.4292563199996948,
      "eval_runtime": 12.4695,
      "eval_samples_per_second": 46.113,
      "eval_steps_per_second": 5.774,
      "step": 12
    },
    {
      "epoch": 2.39,
      "step": 12,
      "total_flos": 7.973531837844685e+16,
      "train_loss": 1.4526023268699646,
      "train_runtime": 1107.1405,
      "train_samples_per_second": 14.006,
      "train_steps_per_second": 0.027
    }
  ],
  "logging_steps": 5,
  "max_steps": 30,
  "num_train_epochs": 3,
  "save_steps": 500,
  "total_flos": 7.973531837844685e+16,
  "trial_name": null,
  "trial_params": null
}