search

mukulpatnaik / researchgpt

A LLM based research assistant that allows you to have a conversation with a research paper
https://www.dara.chat
MIT License
3.55k stars 340 forks source link

save the embedding DataFrame #29

Closed goldengrape closed 1 year ago

goldengrape commented 1 year ago

When running locally, the embedding needs to be calculated every time the same PDF is opened, which is somewhat wasteful. It would be best to save the embedding dataframe so that it can be checked before the next use. If the embedding has already been done, it doesn't need to be done again. Additionally, can the embedding dataframe be used as a search index in Zotero?

mukulpatnaik commented 1 year ago

Hi @goldengrape I have added a way to store embeddings locally using redis. You can find that implementation here. You have to install redis and run redis-serve in a separate terminal for this to work