🦊/Companies

LLaVA

1 known investors

Find your way into LLaVA

LLaVA is an academic research project on visual instruction tuning that connects a CLIP visual encoder with a large language model (Vicuna) to build a multimodal assistant, and releases the LLaVA-Instruct-150K dataset of language-image instruction-following samples. The work targets AI/ML researchers and is licensed for non-commercial research use only.

Founders & leadership

HL
Haotian Liu
CL
Chunyuan Li
YL
Yuheng Li
YJ
Yong Jae Lee
QW
Qingyang Wu

Investors · 1

Frequently asked questions

What does LLaVA do?
LLaVA is an academic research project on visual instruction tuning that connects a CLIP visual encoder with a large language model (Vicuna) to build a multimodal assistant, and releases the LLaVA-Instruct-150K dataset of language-image instruction-following samples. The work targets AI/ML researchers and is licensed for non-commercial research use only.
Who founded LLaVA?
LLaVA was founded by Haotian Liu, Chunyuan Li, Yuheng Li, Yong Jae Lee, Qingyang Wu.
Who are LLaVA's investors?
LLaVA's investors include Gradient Ventures.