trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "global_step": 340,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15,
      "learning_rate": 1.1750000000000001e-05,
      "loss": 15.017,
      "step": 50
    },
    {
      "epoch": 0.29,
      "learning_rate": 2.425e-05,
      "loss": 6.7134,
      "step": 100
    },
    {
      "epoch": 0.44,
      "learning_rate": 3.675e-05,
      "loss": 4.3869,
      "step": 150
    },
    {
      "epoch": 0.59,
      "learning_rate": 4.9250000000000004e-05,
      "loss": 3.6209,
      "step": 200
    },
    {
      "epoch": 0.74,
      "learning_rate": 6.175e-05,
      "loss": 3.2011,
      "step": 250
    },
    {
      "epoch": 0.88,
      "learning_rate": 7.425e-05,
      "loss": 3.0513,
      "step": 300
    }
  ],
  "max_steps": 1700,
  "num_train_epochs": 5,
  "total_flos": 1.7302176965482906e+18,
  "trial_name": null,
  "trial_params": null
}