trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "global_step": 69,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.43,
      "learning_rate": 4.27536231884058e-05,
      "loss": 0.7204,
      "step": 10
    },
    {
      "epoch": 0.87,
      "learning_rate": 3.5507246376811596e-05,
      "loss": 0.6918,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8,
      "eval_loss": 0.6894806623458862,
      "eval_runtime": 0.1526,
      "eval_samples_per_second": 32.764,
      "eval_steps_per_second": 6.553,
      "step": 23
    },
    {
      "epoch": 1.3,
      "learning_rate": 2.826086956521739e-05,
      "loss": 0.6986,
      "step": 30
    },
    {
      "epoch": 1.74,
      "learning_rate": 2.101449275362319e-05,
      "loss": 0.7019,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6,
      "eval_loss": 0.6858540177345276,
      "eval_runtime": 0.1355,
      "eval_samples_per_second": 36.891,
      "eval_steps_per_second": 7.378,
      "step": 46
    },
    {
      "epoch": 2.17,
      "learning_rate": 1.3768115942028985e-05,
      "loss": 0.6927,
      "step": 50
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.521739130434783e-06,
      "loss": 0.69,
      "step": 60
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.6,
      "eval_loss": 0.6853448748588562,
      "eval_runtime": 0.1356,
      "eval_samples_per_second": 36.871,
      "eval_steps_per_second": 7.374,
      "step": 69
    },
    {
      "epoch": 3.0,
      "step": 69,
      "total_flos": 1465242836520960.0,
      "train_loss": 0.6992312652477319,
      "train_runtime": 27.1417,
      "train_samples_per_second": 2.542,
      "train_steps_per_second": 2.542
    }
  ],
  "max_steps": 69,
  "num_train_epochs": 3,
  "total_flos": 1465242836520960.0,
  "trial_name": null,
  "trial_params": null
}