News for hackers. Stuff that matters.

A comment on Dot: use of local LLMs and RAG in particular to interact with documents. Parent comment

Re:Dot: use of local LLMs and RAG in particular to interact ... by simonw on Sunday April 07, 2024 @10:52AM (#39961163)
What options do you think work better?
[ Parent | on HN ]
Re:Dot: use of local LLMs and RAG in particular to interact ... by turnsout on Sunday April 07, 2024 @11:45AM (#39961516)
I don’t have an opinion, just wondering why they didn’t choose other another option such as Sentence Embeddings, OpenAI embeddings, etc.
[ Parent | on HN ]
Re:Dot: use of local LLMs and RAG in particular to interact ... by simonw on Sunday April 07, 2024 @06:34PM (#39964556)
FAISS can be used with OpenAI embeddings (and any other embedding model).

FAISS is technology for fast indexed similarity vector search - using it is an independent decision from which model you use to create those vectors.

[ Parent | on HN ]
Re:Dot: use of local LLMs and RAG in particular to interact ... by J_Shelby_J on Sunday April 07, 2024 @11:02AM (#39961225)
I’m trying build an exhaustive list of realistic options. See the spreadsheet here

https://shelbyjenkins.github.io/blog/retrieval-is-all-you-ne...

[ Parent | on HN ]