FuxiaoLiu / LRV-Instruction

[ICLR'24] Mitigating Hallucination in Large Multi-Modal Models via Robust Instruction Tuning
https://fuxiaoliu.github.io/LRV/
BSD 3-Clause "New" or "Revised" License
254 stars 13 forks source link

About filter_cap.json file #9

Closed yytzsy closed 1 year ago

yytzsy commented 1 year ago

I want to ask about the number of question-answer pairs in filter_cap.json file.

In the paper, it has been stated that there are 120k visual instructions generated by GPT4, covering 16 vision-and-language tasks with openended instructions and answers, however, I can only find about 15k QA pairs in filter_cap.json. Can you provide the full 120k QA-pairs?

yytzsy commented 1 year ago

sorry, I misread the data. I will close this issue