Best french model embedder for retriever LangChain?

#61
by cfrancois7 - opened

I would like to use Mistral with Langchain to do RAG (retrieval augmented generation) but on French document.
I would like to know what is the up to date / best embedder for French language you know for embedding generation and retrieval?

I know this question is not very related to Mistral, but I'm sure the community is up to date about the best french model.
Ask me for english, I'm good, but for french, I know only camenbert models. Is there anything better?

Try this one.
https://huggingface.co/Sahajtomar/french_semantic
Which one is better in English?

Thank you for your reply.
I will test your model and compare with the Javaness Ones : https://huggingface.co/dangvantuan/sentence-camembert-large

Regarding the english, we can just relate to the LLM benchmarks. But for other langages than english, the benchmark is less relevant and consistent.

Sign up or log in to comment