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

# V* Benchmark
refactor [V* Benchmark](https://huggingface.co/datasets/craigwu/vstar_bench) to add support for VLMEvalKit.

# Benchmark Information
1. Number of questions: 191
2. Question type: MCQ (multiple choice question)
3. Question format: image + text

# Reference

- [VLMEvalKit](https://github.com/open-compass/VLMEvalKit)
- [V* Benchmark](https://huggingface.co/datasets/craigwu/vstar_bench)
- [MMVP](https://huggingface.co/datasets/MaoSong2022/MMVP)