3 ms·
Isn‘t clip superseeded by multimodal llms?
by Jack5500 2y ago
Isn‘t clip superseeded by multimodal llms?
- Eisenstein 2y agoIn this program CLIP is being used to create embeddings. A multimodal LLM does something very similar. In this case the language model is not needed because the embeddings are being used to search directly.