checkpoint-2400/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.572347266881029,
  "global_step": 2400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.54,
      "learning_rate": 5e-05,
      "loss": 3.7701,
      "step": 500
    },
    {
      "epoch": 1.07,
      "learning_rate": 3.9125706829056117e-05,
      "loss": 3.4878,
      "step": 1000
    },
    {
      "epoch": 1.61,
      "learning_rate": 2.8251413658112224e-05,
      "loss": 3.1876,
      "step": 1500
    },
    {
      "epoch": 2.14,
      "learning_rate": 1.7377120487168335e-05,
      "loss": 3.0536,
      "step": 2000
    }
  ],
  "max_steps": 2799,
  "num_train_epochs": 3,
  "total_flos": 313419792384000.0,
  "trial_name": null,
  "trial_params": null
}