--- dataset_info: features: - name: image dtype: image - name: bboxes sequence: sequence: int64 - name: text sequence: string - name: language dtype: string splits: - name: train num_bytes: 478482327.362134 num_examples: 4635 download_size: 436363110 dataset_size: 478482327.362134 configs: - config_name: default data_files: - split: train path: data/train-* ---