[Perf] Skip detokenization in online beam search (#46422)

Signed-off-by: Guy Stone <[email protected]>
Signed-off-by: Guy Stone <[email protected]>
Co-authored-by: Claude <[email protected]>
This commit is contained in:
Guy Stone
2026-06-23 15:46:09 -07:00
committed by GitHub
co-authored by Claude
parent 0a3e2dbc09
commit 3cc871aaf1
@@ -61,6 +61,7 @@ class BeamSearchOnlineMixin(ABC):
logprobs=logprobs_num,
max_tokens=1,
temperature=temperature,
detokenize=False,
)
all_beams = [
BeamSearchSequence(