Repository navigation
S26-E7: reject verl with expert parallelism; require fsdp2 strategy - #283
Closed
michaellchung wants to merge 1 commit into
Closed
michaellchung wants to merge 1 commit into
michaellchung wants to merge 1 commit into
Conversation
…ason The verl Megatron engine (the only verl path with EP) is not wired into the yeto adapter; EP runs use the Miles backend. check_expert_parallel now names the decision and the backend to use, and stage1 refuses a verl trainer.verl.strategy other than fsdp2. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Contributor
Author
|
Closing: all commits of |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Decision record for E7: yeto's verl adapter stays FSDP2-only this round (a Megatron backend would rewrite export, hashing, cuts, sampler updates and resume checks; MoE/EP is served by the Miles backend).
Tests: new unit tests; related files 91 passed. The GPU comparison plan (TE vs SequentialMLP under EP8) is on hold.
🤖 Generated with Claude Code