search

mbzuai-oryx / Video-ChatGPT

[ACL 2024 šŸ”„] Video-ChatGPT is a video conversation model capable of generating meaningful conversation about videos. It combines the capabilities of LLMs with a pretrained visual encoder adapted for spatiotemporal video representation. We also introduce a rigorous 'Quantitative Evaluation Benchmarking' for video-based conversational models.
https://mbzuai-oryx.github.io/Video-ChatGPT
Creative Commons Attribution 4.0 International
1.23k stars 108 forks source link

Can I deal with Chinese video? #81

Closed zhangyuereal closed 10 months ago

zhangyuereal commented 10 months ago

Can I deal with Chinese video? Likeļ¼Œask questions and get answer in Chinese?

Need I retrain any model?

mmaaz60 commented 10 months ago

Hi @zhangyuereal,

Our base LLM model is Vicuna which is reasonably good in Chinese. Although Video-ChatGPT is trained using English video conversations, you can ask questions and get response back in Chinese. However, in order to get optimal performance, it is recommended to translate VideoInstruct data to Chinese and retrain the model.

I hope it will be helpful. Thank You.