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": 3.4825870646766175e-05,
      "loss": 1.255,
      "step": 500
    },
    {
      "epoch": 0.75,
      "learning_rate": 1.4096185737976782e-05,
      "loss": 1.0544,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "step": 1340,
      "total_flos": 1.306695716831232e+16,
      "train_loss": 1.1239764996428987,
      "train_runtime": 1102.0635,
      "train_samples_per_second": 38.892,
      "train_steps_per_second": 1.216
    }
  ],
  "max_steps": 1340,
  "num_train_epochs": 1,
  "total_flos": 1.306695716831232e+16,
  "trial_name": null,
  "trial_params": null
}