A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 5, 01:00 UTC
[CI Failure] Fix Plugin Tests (2 GPUs) Failure (#40083)
Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io> Co-authored-by: Michele Gazzetti <michele.gazzetti1@ibm.com>
wang.yuqi committed
Apr 17, 2026 at 12:17 UTC
978a4462bbc529ff204647543526e4caa08ed974
1 file changed
+5
vllm/entrypoints/pooling/pooling/protocol.py
+5
@@ -97,6 +97,11 @@ class IOProcessorRequest(PoolingBasicRequestMixin, EncodingRequestMixin, Generic
97
max_total_tokens_param="max_model_len",
98
)
99
100
+ def to_pooling_params(self):
101
+ return PoolingParams(
102
+ task=self.task,
103
+ )
104
+
105
106
class IOProcessorResponse(OpenAIBaseModel, Generic[T]):
107
request_id: str | None = None