[multi-lora] 1/n - 1, feat: per-token loss-weight/advantage sample channels - #2724
Open
yushengsu-thu wants to merge 2 commits into
Open
[multi-lora] 1/n - 1, feat: per-token loss-weight/advantage sample channels#2724yushengsu-thu wants to merge 2 commits into
yushengsu-thu wants to merge 2 commits into
Conversation
Client-supplied per-token channels ride Sample like the OPD lists: wire dtypes stay float32 while the binary loss_masks remain int32, CP slicing covers them, and multi-turn merges zero them over injected observation spans. AdapterRef gains registration-scoped serving identity fields, and get_reward_value tolerates arg namespaces without reward_key.
yushengsu-thu
marked this pull request as ready for review
August 23, 2026 23:59
yushengsu-thu
requested review from
Shi-Dong,
Zhichenzzz,
fzyzcjy,
guapisolo,
jybsuper,
maocheng23 and
yueming-yuan
as code owners
August 23, 2026 23:59
![claude[bot]](https://avatars.githubusercontent.com/in/1236702?s=60&v=4){kind=small}
There was a problem hiding this comment.
Claude Code Review
This repository is configured for manual code reviews. Comment @claude review for a one-time review, or @claude review always to subscribe this PR to a review on every future push.
Tip: disable this comment in your organization's Code Review settings.
yushengsu-thu
added
run-ci-megatron
run-ci-lora
run-ci-lora-native
This was referenced Aug 24, 2026
yushengsu-thu
removed
run-ci-megatron
run-ci-lora
run-ci-lora-native
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Adds client-supplied per-token channels to the sample data format:
Sample.loss_weights/Sample.advantages(response-aligned likeloss_mask, float32 on the wire while the binaryloss_masksstay int32), CP-slice coverage, and observation-span zero-padding in multi-turn merges.AdapterRefgains registration-scoped serving-identity fields andget_reward_valuetolerates arg namespaces withoutreward_key. All additions are inert until a rollout path populates them.Extracted from #2273; the stack rebases after this merges.