Datasets:
metadata
license: mit
task_categories:
- visual-question-answering
language:
- vi
pretty_name: lavy-bench-gpt4o
size_categories:
- n<1K
LaVy-Bench (with Answers ๐)
Welcome to the LaVy-Bench dataset repository!
About
We offers manually generated answers created using GPT-4, providing meaningful, detailed, and bug-free responses. Our goal is to contribute to LaVy-Bench as a significant benchmark for Vietnamese Multi-Modal and Vietnamese Large Vision Language Models in real-world scenarios.
Contribution
We aim to generate meaningful answers for questions-only datasets sourced from the original LaVy-Bench repository.