--- dataset_info: features: - name: instruction dtype: string - name: predict sequence: 'null' splits: - name: train num_bytes: 1291501 num_examples: 15000 download_size: 659215 dataset_size: 1291501 configs: - config_name: default data_files: - split: train path: data/train-* ---