[BugFix] Fix false assertion with spec-decode=[2,4,..] and TP>2 #29036
Conversation
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
LucasWilkinson
requested review from
ProExpertProg,
WoosukKwon,
hmellor,
houseroad,
mgoin,
robertgshaw2-redhat,
tlrmchlsmth,
yewentao256 and
youkaichao
as code owners
LucasWilkinson
added
the
ready
![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 addresses a bug where an assertion related to sequence parallelism was incorrectly triggered even when sequence parallelism was disabled. The change correctly adds a check for self.pass_config.enable_sequence_parallelism before applying constraints on cudagraph capture sizes related to tensor_parallel_size. This ensures that the logic is only applied when sequence parallelism is active, resolving the false assertion. The fix is correct and well-targeted.
LucasWilkinson
changed the title
LucasWilkinson
force-pushed
the
lwilkinson/mtp-fix
branch
from
100b3fd to
bf20165
Compare
LucasWilkinson
changed the title
mgoin
approved these changes
Nov 19, 2025
mgoin
approved these changes
Nov 19, 2025
khluu
pushed a commit
that referenced
this pull request
Nov 19, 2025
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> (cherry picked from commit 8f4f77a)
Victor49152
pushed a commit
to Victor49152/vllm
that referenced
this pull request
Nov 20, 2025
…-project#29036) Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
LuminolT
pushed a commit
to LuminolT/vllm
that referenced
this pull request
Nov 21, 2025
…-project#29036) Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> Signed-off-by: LuminolT <lumischen01@gmail.com>
bigPYJ1151
pushed a commit
that referenced
this pull request
Nov 25, 2025
Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com> Signed-off-by: jiang1.li <jiang1.li@intel.com>
bringlein
pushed a commit
to bringlein/vllm
that referenced
this pull request
Nov 26, 2025
…-project#29036) Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
devpatelio
pushed a commit
to SumanthRH/vllm
that referenced
this pull request
Nov 29, 2025
…-project#29036) Signed-off-by: Lucas Wilkinson <lwilkins@redhat.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 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.
Fix
vllm serve deepseek-ai/DeepSeek-R1 --speculative-config '{"method": "mtp", "num_speculative_tokens": 2}' -tp 8falsely asserting; should only assert when sequence parallelism is enabled. (Introduced in #28315)Main
PR
Boots fine