Skip to content

[Perf] Skip detokenization in offline beam search #50333

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
vllm-bot merged 1 commit into from
Aug 10, 2026
Merged

[Perf] Skip detokenization in offline beam search #50333

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions vllm/entrypoints/generate/beam_search/offline.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -119,6 +119,7 @@ def beam_search(
logprobs=2 * beam_width,
max_tokens=1,
temperature=temperature,
detokenize=False,
skip_clone=True, # Internal beam search, safe to skip clone
)
instances: list[BeamSearchInstance] = []
Expand Down Expand Up @@ -443,6 +444,7 @@ def _build_beam_sampling_params(
logprobs=base_params.logprobs,
max_tokens=1,
temperature=base_params.temperature,
detokenize=False,
allowed_token_ids=(
allowed_ids
if len(allowed_ids) <= _MAX_NUM_ALLOWED_TOKEN_IDS
Expand Down
Loading