❓ Support openai for Embedding Generation

[nsoybean]

Will there be support for openai for embedding generation in the new future?

[jparkerweb]

@nsoybean

I'm having a hard time understanding the desire to not do the embedding generation close to the code vs 3rd party. Embedding models are cheap, FAST, and small.

Would love to hear why OpenAI is being asked for here.

[edho08]

Embedding model are indeed cheap, FAST and small as you have mentioned. However, they arent cheap or fast enough to run on a small server if we want to provide some kind of RAG service backend. Using 3rd party service might be way more cheaper then say upgrading the server to handle many requests.