Skip ring of experts all gather when EP = 1#4086
Open
Shuwen-Fang wants to merge 1 commit into
Open
Conversation
Shuwen-Fang
requested review from
A9isha,
NicoGrande,
NuojCheng,
RissyRan,
SurbhiJainUSC,
abhinavclemson,
aireenmei,
bvandermoon,
darisoy,
dipannita08,
gagika,
gobbleturk,
hengtaoguo,
igorts-git,
jesselu-google,
jiangjy1982,
khatwanimohit,
michelle-yooh,
richjames0,
shralex,
suexu1025 and
vipannalla
as code owners
Codecov Report✅ All modified and coverable lines are covered by tests. 📢 Thoughts on this report? Let us know! |
NuojCheng
reviewed
Jun 8, 2026
| ici_expert_parallelism: 1 | ||
|
|
||
| # Enabling check_vma is recommended for improved performance. Only supported for EP / FSDP ICI parallelisms, shard_mode: "auto", use_ragged_sort: False, use_ring_of_experts: False, and use_tokamax_gmm=False. | ||
| # Enabling check_vma is recommended for improved performance. Only supported for EP / FSDP ICI parallelisms, shard_mode: "auto", use_ragged_sort: False, and use_tokamax_gmm=False. |
Collaborator
There was a problem hiding this comment.
emmm why this flag is removed?
NuojCheng
reviewed
Jun 8, 2026
| x, logits, pre_bias_logits = tuple( | ||
| jax.lax.all_gather(z, axis_name=self._expert_parallelism_name, tiled=True) | ||
| for z in (x, logits, pre_bias_logits) | ||
| ) |
Collaborator
There was a problem hiding this comment.
I think we should error out if use_ring_of_experts=true and num_ep==1
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Description
Currently runs error out when EP = 1 and ring of experts is used, fix by skipping all gather when EP = 1. This PR updates the logic to only all gather when EP > 1.
Tests
CI tests
Checklist
Before submitting this PR, please make sure (put X in square brackets):
gemini-reviewlabel.