search

haotian-liu / LLaVA

[NeurIPS'23 Oral] Visual Instruction Tuning (LLaVA) built towards GPT-4V level capabilities and beyond.
https://llava.hliu.cc
Apache License 2.0
20.04k stars 2.21k forks source link

多模态语言模型能把LLM单独抽出来使用吗？ #1257

Closed sunnychenxiwang closed 7 months ago

sunnychenxiwang commented 7 months ago

Question

我很好奇，是否能把MLLM的LLM分离出来，还能和之前一样和其文本对话。同时此时的LLM比起之前是否能力下降。

haotian-liu commented 7 months ago

Without the multimodal patches, LLaVA operates as an ordinary LLM. You can try it in our official demo. Thanks.