checkpoint-300/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2,
  "eval_steps": 500,
  "global_step": 300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "learning_rate": 0.0001,
      "loss": 1.96,
      "step": 50
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.0002,
      "loss": 1.0382,
      "step": 100
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00017071067811865476,
      "loss": 1.0275,
      "step": 150
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0001,
      "loss": 1.035,
      "step": 200
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 1.0533,
      "step": 250
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0,
      "loss": 0.986,
      "step": 300
    }
  ],
  "logging_steps": 50,
  "max_steps": 300,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 9315508499251200.0,
  "trial_name": null,
  "trial_params": null
}