trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "global_step": 1340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.37,
      "learning_rate": 6.965174129353234e-07,
      "loss": 1.8401,
      "step": 500
    },
    {
      "epoch": 0.75,
      "learning_rate": 2.8192371475953566e-07,
      "loss": 1.4199,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "step": 1340,
      "total_flos": 1.306756690477056e+16,
      "train_loss": 1.5686970041758979,
      "train_runtime": 1103.7707,
      "train_samples_per_second": 38.833,
      "train_steps_per_second": 1.214
    }
  ],
  "max_steps": 1340,
  "num_train_epochs": 1,
  "total_flos": 1.306756690477056e+16,
  "trial_name": null,
  "trial_params": null
}