dutta18/A-OKVQA-17K

--- dataset_info: features: - name: question dtype: string - name: answer dtype: string - name: image dtype: image - name: reasoning_type dtype: string splits: - name: train num_bytes: 2920036819 num_examples: 17056 - name: validation num_bytes: 188987802 num_examples: 1145 download_size: 3113951848 dataset_size: 3109024621 configs: - config_name: default data_files: - split: train path: data/train-* - split: validation path: data/validation-* --- This is the official AOKVQA dataset with only the train and the val split. And I dont own the copyright of the work. All copyright of this dataset belongs to the original author of the work. Uploaded here beacuse I couldn't find any suitable place to get the data at one click. Now, community members can easily access this dataset and evaluate the models. Original Paper link: https://arxiv.org/abs/2206.01718