DeepSeek-VL2

model
DeepSeek
MoE vision-language model for advanced multimodal understanding.
Paper (arXiv)GitHubHuggingFace

Paper

arXiv: 2412.10302

multimodalopen-weightmoe

Related