NVIDIA / NVIDIA/Megatron-LM

[QUESTION] MoE score for aux loss

Open
#2,767 3 comments 0 reactions 1 assignee Claimed by @Connor-XY View on GitHub
community-request module: moe question waiting-on-customer
Dominant language
Python
Stars
17.9k
Forks
4.5k
Avg merge
4d 6h
Merged PRs (30d)
271

Description

I noticed that compute_routing_scores_for_aux_loss in megatron/core/transformer/moe/moe_utils.py does not handle grouped topk routing, while topk_routing_with_score_function does. This may cause the auxiliary loss to optimize for a different routing distribution than what is actually used for token dispatching.

Is this intentional behavior, or is it a bug?
If intentional, what is the reasoning behind using different routing logic for aux loss vs. actual token dispatching?

Contributor guide

Open the contributing guide

Assessment

This issue has not been assessed yet.

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.