[Frontend][1/n] Make pooling entrypoints request schema consensus | CompletionRequest (#32395)

Signed-off-by: wang.yuqi <yuqi.wang@daocloud.io>
This commit is contained in:
wang.yuqi
2026-01-16 06:17:04 +00:00
committed by GitHub
parent 73f635a75f
commit 4ae77dfd42
22 changed files with 635 additions and 600 deletions
+2 -10
View File
@@ -10,6 +10,7 @@ from pydantic import (
from vllm import PoolingParams
from vllm.config.pooler import get_use_activation
from vllm.entrypoints.openai.engine.protocol import OpenAIBaseModel, UsageInfo
from vllm.entrypoints.pooling.base.protocol import PoolingBasicRequestMixin
from vllm.entrypoints.pooling.embed.protocol import (
EmbeddingChatRequest,
EmbeddingCompletionRequest,
@@ -72,17 +73,8 @@ class PoolingChatRequest(EmbeddingChatRequest):
T = TypeVar("T")
class IOProcessorRequest(OpenAIBaseModel, Generic[T]):
model: str | None = None
priority: int = Field(default=0)
"""
The priority of the request (lower means earlier handling;
default: 0). Any priority other than 0 will raise an error
if the served model does not use priority scheduling.
"""
class IOProcessorRequest(PoolingBasicRequestMixin, Generic[T]):
data: T
task: PoolingTask = "plugin"
encoding_format: EncodingFormat = "float"
embed_dtype: EmbedDType = Field(