Datasets:
metadata
task_categories:
- question-answering
- visual-question-answering
- image-to-text
language:
- en
size_categories:
- 1K<n<10K