[Deprecation] Remove scheduled deprecated items (#60923)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
This commit is contained in:
@@ -4,7 +4,6 @@ from collections.abc import Sequence
|
||||
from typing import Any
|
||||
|
||||
from vllm import PoolingParams, PoolingRequestOutput
|
||||
from vllm.logger import init_logger
|
||||
from vllm.plugins.io_processors import get_io_processor
|
||||
from vllm.renderers.inputs.preprocess import parse_model_prompt, prompt_to_seq
|
||||
|
||||
@@ -21,8 +20,6 @@ from ..typing import (
|
||||
)
|
||||
from .protocol import IOProcessorRequest, IOProcessorResponse
|
||||
|
||||
logger = init_logger(__name__)
|
||||
|
||||
|
||||
class UnsupportedCombinedTaskIOProcessor(PoolingIOProcessor):
|
||||
name = "embed&token_classify"
|
||||
@@ -117,23 +114,7 @@ class PluginWithIOProcessorPlugins(PoolingIOProcessor):
|
||||
ctx.final_res_batch,
|
||||
request_id=ctx.request_id,
|
||||
)
|
||||
|
||||
if callable(
|
||||
output_to_response := getattr(self.io_processor, "output_to_response", None)
|
||||
):
|
||||
logger.warning_once(
|
||||
"`IOProcessor.output_to_response` is deprecated. To ensure "
|
||||
"consistency between offline and online APIs, "
|
||||
"`IOProcessorResponse` will become a transparent wrapper "
|
||||
"around output data from v0.19 onwards.",
|
||||
)
|
||||
|
||||
if hasattr(output, "request_id") and output.request_id is None:
|
||||
output.request_id = ctx.request_id # type: ignore
|
||||
|
||||
ctx.response = output_to_response(output) # type: ignore
|
||||
else:
|
||||
ctx.response = IOProcessorResponse(request_id=ctx.request_id, data=output)
|
||||
ctx.response = IOProcessorResponse(request_id=ctx.request_id, data=output)
|
||||
|
||||
#######################################
|
||||
# offline APIs
|
||||
|
||||
@@ -32,7 +32,6 @@ router = APIRouter()
|
||||
async def pause_generation(
|
||||
raw_request: Request,
|
||||
mode: Annotated[PauseMode, Query()] = "abort",
|
||||
wait_for_inflight_requests: bool = Query(False),
|
||||
clear_cache: Annotated[bool, Query()] = True,
|
||||
) -> JSONResponse:
|
||||
"""Pause generation requests to allow weight updates.
|
||||
@@ -44,7 +43,6 @@ async def pause_generation(
|
||||
- ``"abort"``: Abort all in-flight requests immediately (default).
|
||||
- ``"wait"``: Wait for in-flight requests to complete.
|
||||
- ``"keep"``: Freeze requests in queue; they resume on /resume.
|
||||
wait_for_inflight_requests: DEPRECATED. Use ``mode="wait"`` instead.
|
||||
clear_cache: DEPRECATED. Whether to clear KV/prefix caches after
|
||||
draining. Ignored when mode="keep".
|
||||
|
||||
@@ -55,7 +53,6 @@ async def pause_generation(
|
||||
await engine.pause_generation(
|
||||
mode=mode,
|
||||
clear_cache=clear_cache,
|
||||
wait_for_inflight_requests=wait_for_inflight_requests,
|
||||
)
|
||||
return JSONResponse(
|
||||
content={"status": "paused"},
|
||||
|
||||
Reference in New Issue
Block a user