[DRAFT] added support for 3d weight params arising from conv1d #69
Conversation
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
Signed-off-by: mikail <[email protected]>
skyw
reviewed
Nov 3, 2025
| if partition_dim is None: | ||
| # Fallback path for non TP params. | ||
| # Handle 3D conv1d case | ||
| if x.dim() == 3: |
There was a problem hiding this comment.
This is not the right place to add this logic.
the function still handles 2d input, reshape logic should be outside of this function. probably should be in the Megatron inherited muon
There was a problem hiding this comment.
The issue is the same as unfusing QKV: in the optimizer state, there param shape will always be 3D. We would need to add a handle in the OrthogonalizedOptimizer class
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.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.
No description provided.