Datasets:
metadata
task_categories:
- image-text-to-text
This repository contains the data for Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources.
Project page: https://victorwz.github.io/Open-Qwen2VL