Recommendations model quantisation #6
Merged
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
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Outdated suggestions cannot be applied.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.
Closes #5
By default, model quantisation (int4) is turned off. (Results are much worse and avoids a breaking change.)
The Quanto integration in the Transformers library is fairly new, and it's broken in the few releases that have it. So for now a temporary build from main was created at ae9is/transformers that fixes it. (This seemed simpler than setting up the Docker builds to checkout and build the source repository.)
With int4 quantisation the model api needs only ~256M instead of ~512M, which was right around the cut-off for small VMs and causing out of memory errors.