Fix repetition penalty crash in decode phase #769
Merged
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
pawel-olejniczak
requested review from
adobrzyn,
afierka-intel,
iboiko-habana,
kamil-kaczor,
ksmusz,
kzawora-intel,
mgawarkiewicz-intel,
michalkuligowski and
xuechendi
as code owners
The sampler crashes when repetition penalties are used because make_selective_sampling_metadata() sets prompt_token_ids=None during skip_copy=True (decode phase), but penalties require prompt tokens. Add caching mechanism for prompt_token_ids to reuse the tensor when skip_copy=True and penalties are needed. Cache is invalidated when batch composition changes. Fixes repetition penalty support while preserving skip_copy performance optimization. Signed-off-by: Paweł Olejniczak <[email protected]>
pawel-olejniczak
force-pushed
the
dev/polejnix/fix_repetition_penalty_crash
branch
from
d8fe456 to
b76d0a0
Compare
✅ CI PassedAll checks passed successfully against the following vllm commit: |
iboiko-habana
approved these changes
Jan 5, 2026
yingjie-han
pushed a commit
to yingjie-han/vllm-gaudi
that referenced
this pull request
Jan 16, 2026
The sampler crashes when repetition penalties are used because make_selective_sampling_metadata() sets prompt_token_ids=None during skip_copy=True (decode phase), but penalties require prompt tokens. Add caching mechanism for prompt_token_ids to reuse the tensor when skip_copy=True and penalties are needed. Cache is invalidated when batch composition changes. Fixes repetition penalty support while preserving skip_copy performance optimization. Signed-off-by: Paweł Olejniczak <[email protected]>
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.
The sampler crashes when repetition penalties are used because make_selective_sampling_metadata() sets prompt_token_ids=None during skip_copy=True (decode phase), but penalties require prompt tokens. Add caching mechanism for prompt_token_ids to reuse the tensor when skip_copy=True and penalties are needed. Cache is invalidated when batch composition changes. Fixes repetition penalty support while preserving skip_copy performance optimization.