# LVIS-Instruct4V-Nodetail-mix619k

This is a mixture of our LVIS-Instruct4V dataset (no detail and only conversation) with the academic task related data, including ShareGPT, VQAv2, GQA, OKVQA, OCRVQA, AOKVQA, TextCaps, RefCOCO, and VG.

Please refer to the Sec. 3 and pseudo-code in [our paper](https://arxiv.org/abs/2311.07574) for the clarification of ***detail*** and ***conversation*** in our dataset.