biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

add io_process_plugin for sparse embedding (#34214)

Browse Source 
Signed-off-by: augusto.yjh <augusto.yjh@antgroup.com>
Signed-off-by: Augusto Yao <augusto.yjh@antgroup.com>
Co-authored-by: Cyrus Leung <cyrus.tl.leung@gmail.com>
This commit is contained in:
Augusto Yao
2026-02-28 17:16:37 +08:00
committed by GitHub
parent 0892d1ab1f
commit 8e75d88554
14 changed files with 441 additions and 10 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
docs/design/io_processor_plugins.md
View File

@@ -13,12 +13,13 @@ IOProcessorInput = TypeVar("IOProcessorInput")
IOProcessorOutput = TypeVar("IOProcessorOutput")
class IOProcessor(ABC, Generic[IOProcessorInput, IOProcessorOutput]):
def __init__(self, vllm_config: VllmConfig):
"""Abstract interface for pre/post-processing of engine I/O."""
def __init__(self, vllm_config: VllmConfig, renderer: BaseRenderer):
super().__init__()
self.vllm_config = vllm_config
@abstractmethod
def parse_data(self, data: object) -> IOProcessorInput:
raise NotImplementedError
@@ -32,7 +33,7 @@ class IOProcessor(ABC, Generic[IOProcessorInput, IOProcessorOutput]):
self,
params: PoolingParams | None = None,
) -> PoolingParams:
return params or PoolingParams()
return params or PoolingParams(task="plugin")
@abstractmethod
def pre_process(