xinyun/vllm
1
0
Fork 0
mirror of https://git.datalinker.icu/vllm-project/vllm.git synced 2026-07-09 15:47:20 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix dp router

Browse Source 
Signed-off-by: inkcherry <mingzhi.liu@amd.com>
This commit is contained in:
inkcherry 2025-11-19 08:35:28 +00:00
parent a7ea23d16d
commit 675943e018
2 changed files with 3 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
vllm/entrypoints/openai/serving_completion.py
View File

@ -225,6 +225,7 @@ class OpenAIServingCompletion(OpenAIServing):
lora_request=lora_request, lora_request=lora_request,
trace_headers=trace_headers, trace_headers=trace_headers,
priority=request.priority, priority=request.priority,
data_parallel_rank=data_parallel_rank,
) )
generator = self.engine_client.generate( generator = self.engine_client.generate(

2
vllm/entrypoints/openai/serving_engine.py
View File

@ -1172,6 +1172,7 @@ class OpenAIServing:
lora_request: LoRARequest | None, lora_request: LoRARequest | None,
trace_headers: Mapping[str, str] | None, trace_headers: Mapping[str, str] | None,
priority: int, priority: int,
data_parallel_rank: int,
) -> tuple[EngineCoreRequest, dict[str, Any]]: ) -> tuple[EngineCoreRequest, dict[str, Any]]:
"""Use the Processor to process inputs for AsyncLLM.""" """Use the Processor to process inputs for AsyncLLM."""
tokenization_kwargs: dict[str, Any] = {} tokenization_kwargs: dict[str, Any] = {}
@ -1187,6 +1188,7 @@ class OpenAIServing:
tokenization_kwargs=tokenization_kwargs, tokenization_kwargs=tokenization_kwargs,
trace_headers=trace_headers, trace_headers=trace_headers,
priority=priority, priority=priority,
data_parallel_rank=data_parallel_rank,
) )
return engine_request, tokenization_kwargs return engine_request, tokenization_kwargs