trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.0,
  "global_step": 76,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.53,
      "learning_rate": 0.0001736842105263158,
      "loss": 0.8297,
      "step": 10
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014736842105263158,
      "loss": 0.7318,
      "step": 20
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012105263157894738,
      "loss": 0.5737,
      "step": 30
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.473684210526316e-05,
      "loss": 0.7944,
      "step": 40
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.842105263157895e-05,
      "loss": 0.5346,
      "step": 50
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.210526315789474e-05,
      "loss": 0.5529,
      "step": 60
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.5789473684210526e-05,
      "loss": 0.4213,
      "step": 70
    },
    {
      "epoch": 4.0,
      "step": 76,
      "total_flos": 9.175133803452826e+16,
      "train_loss": 0.6096237082230417,
      "train_runtime": 230.614,
      "train_samples_per_second": 5.134,
      "train_steps_per_second": 0.33
    }
  ],
  "max_steps": 76,
  "num_train_epochs": 4,
  "total_flos": 9.175133803452826e+16,
  "trial_name": null,
  "trial_params": null
}