[sglang-miles] R3 (Rollout Routing Replay) DeepEP and MTP support - #18642
Merged
yueming-yuan merged 1 commit intoFeb 12, 2026
Merged
Conversation
yueming-yuan
requested review from
AniZpZ,
BBuf,
CatherineSue,
Edwardf0t1,
FlamingoPg,
Fridge003,
HaiShaw,
JustinTong0323,
Ying1123,
ch-wan,
hnyls2002,
ispobock,
merrymercy,
slin1237 and
xiezhq-hermann
as code owners
February 11, 2026 23:22
Contributor
|
Warning You have reached your daily quota limit. Please wait up to 24 hours and I will start processing your requests again! |
…ve decoding - routed_experts_capturer: Add DeepEP all_gather support for expert routing capture across TP ranks, skip DP slicing when DeepEP handles gathering - model_runner: Skip routed_experts_capturer init/sync for draft workers in speculative decoding, fix cuda_graph_num_tokens to account for num_tokens_per_bs in speculative decoding - scheduler_output_processor_mixin: Fix detokenizer send condition to use rids instead of reqs to ensure streaming outputs are delivered Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
yueming-yuan
force-pushed
the
sglang-miles-r3
branch
from
February 11, 2026 23:24
72352a7 to
8d5db4c
Compare
yueming-yuan
added a commit
that referenced
this pull request
Feb 13, 2026
…8642) Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
zianglih
pushed a commit
to zianglih/sglang
that referenced
this pull request
Feb 28, 2026
…l-project#18642) Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
yueming-yuan
added a commit
that referenced
this pull request
Mar 5, 2026
yueming-yuan
added a commit
that referenced
this pull request
Mar 27, 2026
yueming-yuan
added a commit
that referenced
this pull request
Apr 6, 2026
yueming-yuan
added a commit
that referenced
this pull request
May 21, 2026
JessicaJiang-123
pushed a commit
to JessicaJiang-123/sglang
that referenced
this pull request
Jun 21, 2026
nanjiangwill
pushed a commit
to nanjiangwill/sglang
that referenced
this pull request
Jul 7, 2026
XinyuJiangCMU
pushed a commit
to XinyuJiangCMU/sglang
that referenced
this pull request
Jul 17, 2026
yueming-yuan
added a commit
that referenced
this pull request
Jul 25, 2026
Arist12
pushed a commit
to Arist12/sglang
that referenced
this pull request
Aug 27, 2026
Kh4L
pushed a commit
to Kh4L/sglang
that referenced
this pull request
Sep 8, 2026
…ort (sgl-project#18642) (cherry picked from commit f78421f)
46 tasks
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.
Summary
routed_experts_capturer.py, enabling correct R3 routing data collection across TP ranks when using DeepEP expert parallelismrouted_experts_capturerinitialization for draft workers and fixcuda_graph_num_tokenscalculation for speculative decoding inmodel_runner.pyreqs→ridsbug inscheduler_output_processor_mixin.pyto ensure streaming outputs are delivered to detokenizer