[Frontend] Split ServingRender into renderer and entrypoint. (#44285)

Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
wang.yuqi
2026-06-23 11:19:09 +00:00
committed by GitHub
parent d32575a2d2
commit 31ca9504b1
26 changed files with 1155 additions and 939 deletions
+3 -3
View File
@@ -19,7 +19,7 @@ from vllm.entrypoints.serve.disagg.protocol import (
DerenderCompletionRequest,
GenerateRequest,
)
from vllm.entrypoints.serve.render.serving import OpenAIServingRender
from vllm.entrypoints.serve.render.serving import ServingRender
from vllm.entrypoints.serve.utils.api_utils import validate_json_request
from vllm.logger import init_logger
@@ -28,8 +28,8 @@ logger = init_logger(__name__)
router = APIRouter()
def render(request: Request) -> OpenAIServingRender | None:
return getattr(request.app.state, "openai_serving_render", None)
def render(request: Request) -> ServingRender | None:
return getattr(request.app.state, "serving_render", None)
@router.post(