fix: multiply H by eta in manifold_muon#8
Open
Jah-yee wants to merge 1 commit into
Open
Conversation
This addresses the issue where the eta learning rate was not being applied to the tangent constraint metric H calculation.
Author
|
Good day, I wanted to drop by and encourage you on this PR! The fix looks clean and addresses the eta multiplication issue precisely. Having the tangent constraint metric correctly scaled by the learning rate is important for the manifold optimization. This PR seems ready to land - the fix is straightforward and the motivation is clear from issue #1. Hope to see it merged soon! Thank you for your work on this excellent library! Warmly, |
This file contains hidden or 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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Good day,
I noticed that in issue #1, there was a comment about the code forgetting the multiplication with eta when assembling H.
This PR fixes the issue by properly multiplying H by eta in the manifold_muon function, ensuring the tangent constraint metric is correctly scaled by the learning rate.
Changes
This addresses the concern raised in issue #1 about the missing eta multiplication in the H calculation.
Thank you for your work on this excellent library!
Warmly,
RoomWithOutRoof