[Frontend] Update OpenAI error response to upstream format (#22099)

Signed-off-by: Moritz Sanft <58110325+msanft@users.noreply.github.com>
This commit is contained in:
Moritz Sanft
2025-08-06 23:06:00 -07:00
committed by GitHub
parent cbc8457b26
commit 370661856b
10 changed files with 73 additions and 67 deletions
+9 -12
View File
@@ -47,10 +47,10 @@ from vllm.entrypoints.openai.protocol import (ChatCompletionRequest,
EmbeddingChatRequest,
EmbeddingCompletionRequest,
EmbeddingRequest,
EmbeddingResponse, ErrorResponse,
PoolingResponse, RerankRequest,
ResponsesRequest, ScoreRequest,
ScoreResponse,
EmbeddingResponse, ErrorInfo,
ErrorResponse, PoolingResponse,
RerankRequest, ResponsesRequest,
ScoreRequest, ScoreResponse,
TokenizeChatRequest,
TokenizeCompletionRequest,
TokenizeResponse,
@@ -412,21 +412,18 @@ class OpenAIServing:
message: str,
err_type: str = "BadRequestError",
status_code: HTTPStatus = HTTPStatus.BAD_REQUEST) -> ErrorResponse:
return ErrorResponse(message=message,
type=err_type,
code=status_code.value)
return ErrorResponse(error=ErrorInfo(
message=message, type=err_type, code=status_code.value))
def create_streaming_error_response(
self,
message: str,
err_type: str = "BadRequestError",
status_code: HTTPStatus = HTTPStatus.BAD_REQUEST) -> str:
json_str = json.dumps({
"error":
json_str = json.dumps(
self.create_error_response(message=message,
err_type=err_type,
status_code=status_code).model_dump()
})
status_code=status_code).model_dump())
return json_str
async def _check_model(
@@ -445,7 +442,7 @@ class OpenAIServing:
if isinstance(load_result, LoRARequest):
return None
if isinstance(load_result, ErrorResponse) and \
load_result.code == HTTPStatus.BAD_REQUEST.value:
load_result.error.code == HTTPStatus.BAD_REQUEST.value:
error_response = load_result
return error_response or self.create_error_response(