You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Just wanted to add, that for some languages, Anthropic as LLM (i.e. Sonnet 3.5) + Cohere as embedding model gives much superior results to either OpenAI or any other Open Source embedding model.
I don't know how much effort it woud be to integrate their API's, but this would be a dramatic improvement for non-English texts. The multilingual Cohere Embeddings make it possible to search related documents among several languages at once.
The text was updated successfully, but these errors were encountered:
Just wanted to add, that for some languages, Anthropic as LLM (i.e. Sonnet 3.5) + Cohere as embedding model gives much superior results to either OpenAI or any other Open Source embedding model.
I don't know how much effort it woud be to integrate their API's, but this would be a dramatic improvement for non-English texts. The multilingual Cohere Embeddings make it possible to search related documents among several languages at once.
The text was updated successfully, but these errors were encountered: