---
task_categories:
- multiple-choice
language:
- en
size_categories:
- n<1K
---

# MMVP Benchmark
refactor [MMVP](https://huggingface.co/datasets/MMVP/MMVP) to support VLMEalKit

# Benchmark Information
1. number of questions: 300
2. question type: multiple choice question
3. question format: image + text

# Reference

- [VLMEvalKit](https://github.com/open-compass/VLMEvalKit)
- [MMVP](https://huggingface.co/datasets/MMVP/MMVP)