Fix: LazyTurtle materialization for non-square fused experts#2707
Merged
Fix: LazyTurtle materialization for non-square fused experts#2707
Conversation
…final tensor layout from the target leaf shape instead of relying on a hard-coded split dimension Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
This PR fixes LazyTurtle rematerialization for fused MoE expert checkpoints that do not match the previously hard-coded split layout.
The original logic could incorrectly materialize tensors such as
gate_proj.weightby assuming a fixed split dimension. That breaks models with rectangular expert projections and models that store fused expert weights intransposed layouts.