question about using image/text embedding separately

OpenBMB / MiniCPM-V

MiniCPM-Llama3-V 2.5: A GPT-4V Level Multimodal LLM on Your Phone

Apache License 2.0

7.86k stars 547 forks source link

Closed zhwuwuwu closed 4 weeks ago

zhwuwuwu commented 1 month ago

可以分别单独使用模型的image和text embedding吗？想用embedding结果做retrival。

bokesyo commented 1 month ago

你好，这个模型是language modeling，不是CLIP所以不能做检索