[MoE] Fix misuse of num_experts as expert parallel group size (ep_size) #7551
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Outdated suggestions cannot be applied.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.
Fixes #7535
Description
This PR fixes a bug in inference/engine.py where num_experts (moe_experts) was incorrectly passed as the expert parallel group size (ep_size) when creating expert parallel groups.
Currently:
This causes invalid behavior whenever
num_experts > world_size, because_create_ep_parallel_groupexpects a group size, not the total number of experts as pointed out by @ArnoochkaRoot Cause
num_experts = number of experts inside the MoE layer.
ep_size = how many GPUs to group together for expert parallelism.
These were mixed up in the code.
##Fix
Replaced the incorrect call with the proper ep_size argument:
Additionally, added a safety check in _create_ep_parallel_group to catch invalid configurations:
Backward compatibility