pubmed_train_processed

--- configs: - config_name: default data_files: - split: train path: data/train-* dataset_info: features: - name: id dtype: string - name: conversations list: - name: from