--- dataset_info: features: - name: audio dtype: audio - name: turno_id dtype: int64 - name: duration dtype: string - name: text dtype: string splits: - name: train num_bytes: 3968057413.209 num_examples: 3089 download_size: 3873463345 dataset_size: 3968057413.209 configs: - config_name: default data_files: - split: train path: data/train-* ---