dataset_info: features: - name: image dtype: image - name: output_json_dumpsed dtype: string configs: - config_name: default data_files: - split: train path: data/*.parquet language: - en
