Visual Instruction Tuning (LLaVA)

Project page:

• https://llava-vl.github.io/

• Similar to PaliGemma, they also use CLIP as the visual encoder

Follow up paper:

• https://arxiv.org/abs/2310.03744