Datasets:
Tasks:
Question Answering
Modalities:
Image
Formats:
imagefolder
Languages:
English
Size:
1K - 10K
License:
Update dataset card: Add primary paper link, refine task categories, and add tags
#2
by
nielsr
HF Staff
- opened
This PR updates the dataset card for MV-MATH to provide more accurate and comprehensive information.
Key changes include:
- Correct Paper Link: Replaced the general survey paper link with the direct link to the MV-MATH dataset's original paper: MV-MATH: Evaluating Multimodal Math Reasoning in Multi-Visual Contexts.
- Refined Task Categories: Added
image-text-to-textto thetask_categoriesmetadata, alongsidequestion-answering, to better reflect the dataset's multimodal input and output nature. - Improved Tags: Added relevant tags (
mathematics,multimodal,reasoning,geometry,education) to enhance the dataset's discoverability on the Hugging Face Hub.
These updates ensure the dataset card accurately represents MV-MATH and helps users find it more easily.