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 16, 00:45 UTC
[Core] Make raw_request optional in ServingCompletion (#12503)
Signed-off-by: Sebastian Schönnenbeck <sebastian.schoennenbeck@comma-soft.com>
Sebastian Schoennenbeck committed
Jan 28, 2025 at 11:56 UTC
2079e43beecc486a607c9d79ab691e0e4563aa11
1 file changed
+2
-2
vllm/entrypoints/openai/serving_completion.py
+2
-2
@@ -58,7 +58,7 @@ class OpenAIServingCompletion(OpenAIServing):
58
async def create_completion(
59
self,
60
request: CompletionRequest,
61
- raw_request: Request,
61
+ raw_request: Optional[Request] = None,
62
) -> Union[AsyncGenerator[str, None], CompletionResponse, ErrorResponse]:
63
"""Completion API similar to OpenAI's API.
64
@@ -137,7 +137,7 @@ class OpenAIServingCompletion(OpenAIServing):
137
lora_request=lora_request,
138
prompt_adapter_request=prompt_adapter_request)
139
140
- trace_headers = (await
140
+ trace_headers = (None if raw_request is None else await
141
self._get_trace_headers(raw_request.headers))
142
143
if isinstance(sampling_params, BeamSearchParams):