Add no_avx2 to FAISS vector store #4890
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
lgtm, thanks @AliMirlou! |
|
fixed lint and merged in #4895! |
|
@dev2049 Too bad though that my name didn't make it to the releases 😅 |
|
Can I ask to become a collaborator? I'm deeply involved with LangChain now and I like to contribute more. |
|
We're really trying to keep the collaborator team small and tight-knit at the moment, so we're only considering folks who have made a good number of contributions. Definitely open to it as time goes on, but want to work with you some more first. But we really do appreciate contributions from anyone and everyone and will do our best to review asap! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Add no_avx2 to FAISS vector store
This PR adds the
no_avx2argument andFAISS_NO_AVX2environment variable todependable_faiss_importfunction to not load FAISS vector store with AVX2 optimization.This option comes in handy when the developer wants to save the vector store to a file and load it on somewhere else. Most modern CPUs have AVX2 optimization but there are situations where the hardware is old, the
faiss-cpupypi package is not compiled with AVX2 flag or this feature is disabled for some reason.I encountered this issue on a PaaS where FAISS pickle and database files created by
save_indexin Colab could not be loaded inside the Docker containers due to the platform not supporting AVX2.@dev2049