llava-hf

/

vip-llava-7b-hf

like 12

Vip is short for visual prompts, as this is a specific llava model trained on images containing visual prompts like boxes and scribbles.

Thanks for the quick reply and answer! Is there any paper or blog post about this method and data?

Yes, I've updated the model card. Here is the project page: https://vip-llava.github.io/