[Spec] Polish FutureMap after #25879: rename callback, async guard, cleanup#25962
Merged
Conversation
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
Contributor
There was a problem hiding this comment.
Code Review
This pull request refines the overlap scheduling and speculative decoding logic to improve robustness and device-agnosticism. Key updates include replacing the sentinel value for seq_lens with None to ensure safer batch filtering and merging, and renaming on_verify_complete to on_publish for better clarity. Additionally, the publish_ready event is now initialized lazily to support different hardware backends, and the stash method includes a guard for empty indices to handle data-parallel idle ranks. I have no feedback to provide.
hnyls2002
requested review from
ByronHsu,
Fridge003,
ShangmingCai and
ispobock
as code owners
…lish # Conflicts: # python/sglang/srt/managers/overlap_utils.py # python/sglang/srt/managers/scheduler.py # python/sglang/srt/speculative/eagle_worker_v2.py # python/sglang/srt/speculative/multi_layer_eagle_worker_v2.py
hnyls2002
added
high priority
and removed
high priority
run-ci
bypass-fastfail
run-ci-extra
labels
hnyls2002
changed the title
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Stacked on #25879. Strictly polish — no behavior change.
on_verify_completekwarg →on_publish(disambiguate fromon_verify_complete_cpumethod)FutureMap.stashpayload asUnion[torch.Tensor, EagleDraftInput]if batch.is_spec_v2gate aroundresolve_seq_lens_cpu; let it self-gate onspec_info.future_indicespublish_ready(lazy viaget_device_module(device).Event(), hip/cuda-agnostic)# DP idlecomment instashto keep the original 'why early-return' reasoning(batch.seq_lens > 0).all()afterresolve_future: catches a-future_indices.indicessentinel slipping through ifpublish_readyfencing or buf indexing is wrong