[Model Runner V2] Bug fix: logprob dtype int64/int32 issue (#41761)

Signed-off-by: yewentao256 <zhyanwentao@126.com>
Signed-off-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: Nick Hill <nickhill123@gmail.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
This commit is contained in:
Wentao Ye
2026-05-11 21:55:43 +00:00
committed by GitHub
co-authored by Nick Hill mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
parent bbee532988
commit d7af6b34d8
3 changed files with 83 additions and 7 deletions
+8
View File
@@ -690,6 +690,14 @@ class SamplingParams(
parameter="logprob_token_ids",
value=n,
)
if self.logprobs is not None and self.logprobs != n:
raise VLLMValidationError(
f"When both logprobs and logprob_token_ids are set, "
f"logprobs must equal len(logprob_token_ids). Got "
f"logprobs={self.logprobs}, len(logprob_token_ids)={n}.",
parameter="logprob_token_ids",
value=n,
)
# Validate prompt logprobs.
if num_prompt_logprobs := self.prompt_logprobs: