trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.4691358024691357,
  "global_step": 2600,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.47,
      "learning_rate": 2.8575498575498578e-05,
      "loss": 1.0064,
      "step": 500
    },
    {
      "epoch": 0.95,
      "learning_rate": 2.715099715099715e-05,
      "loss": 0.7464,
      "step": 1000
    },
    {
      "epoch": 1.42,
      "learning_rate": 2.572649572649573e-05,
      "loss": 0.586,
      "step": 1500
    },
    {
      "epoch": 1.9,
      "learning_rate": 2.4301994301994302e-05,
      "loss": 0.5616,
      "step": 2000
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.287749287749288e-05,
      "loss": 0.4552,
      "step": 2500
    }
  ],
  "max_steps": 10530,
  "num_train_epochs": 10,
  "total_flos": 249833139915718656,
  "trial_name": null,
  "trial_params": null
}