fix: propagate moe_token_dispatcher_type in bridge model provider#1737
Merged
zhuzilin merged 1 commit intoTHUDM:mainfrom Mar 22, 2026
Merged
Conversation
Bridge provider was missing moe_token_dispatcher_type propagation. Since slime always sets variable_seq_lengths=True and allgather dispatcher is incompatible with it, finalize() raises a ValueError. Propagate moe_token_dispatcher_type from args (already corrected to alltoall by validate_args) when the attribute exists. For newer Megatron versions where the attribute was removed from args, the provider is left untouched. Closes THUDM#1725 Made-with: Cursor
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
fix #1725
moe_token_dispatcher_typefromargsto the bridge provider inmodel_provider.py, guarded byhasattrfor compatibility with newer Megatron versions where the attribute may no longer exist onargs.ValueError: Token dispatcher type: allgather does not support variable sequence lengthraised byfinalize()when using bridge mode, since slime always setsvariable_seq_lengths=True.