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

[V0 Deprecation] Remove LLMEngine (#25033)

Browse Source 
Signed-off-by: Woosuk Kwon <woosuk@thinkingmachines.ai>
Signed-off-by: Woosuk Kwon <woosuk.kwon@berkeley.edu>
This commit is contained in:
Woosuk Kwon
2025-09-20 17:56:30 -07:00
committed by GitHub
parent 367a480bd3
commit 52c2a8d4ad
29 changed files with 65 additions and 2763 deletions
Download Patch File Download Diff File Expand all files Collapse all files

37
tests/plugins_tests/test_scheduler_plugins.py
View File

@@ -3,47 +3,18 @@
import pytest
from vllm.core.scheduler import Scheduler
from vllm.engine.arg_utils import EngineArgs
from vllm.engine.llm_engine import LLMEngine
from vllm.sampling_params import SamplingParams
from vllm.v1.core.sched.scheduler import Scheduler as V1Scheduler
from vllm.v1.engine.llm_engine import LLMEngine as V1LLMEngine
from vllm.v1.core.sched.scheduler import Scheduler
from vllm.v1.engine.llm_engine import LLMEngine
class DummyV0Scheduler(Scheduler):
def schedule(self):
raise Exception("Exception raised by DummyV0Scheduler")
class DummyV1Scheduler(V1Scheduler):
class DummyV1Scheduler(Scheduler):
def schedule(self):
raise Exception("Exception raised by DummyV1Scheduler")
def test_scheduler_plugins_v0(monkeypatch: pytest.MonkeyPatch):
with monkeypatch.context() as m:
m.setenv("VLLM_USE_V1", "0")
with pytest.raises(Exception) as exception_info:
engine_args = EngineArgs(
model="facebook/opt-125m",
enforce_eager=True, # reduce test time
scheduler_cls=DummyV0Scheduler,
)
engine = LLMEngine.from_engine_args(engine_args=engine_args)
sampling_params = SamplingParams(max_tokens=1)
engine.add_request("0", "foo", sampling_params)
engine.step()
assert str(
exception_info.value) == "Exception raised by DummyV0Scheduler"
def test_scheduler_plugins_v1(monkeypatch: pytest.MonkeyPatch):
with monkeypatch.context() as m:
m.setenv("VLLM_USE_V1", "1")
@@ -59,7 +30,7 @@ def test_scheduler_plugins_v1(monkeypatch: pytest.MonkeyPatch):
scheduler_cls=DummyV1Scheduler,
)
engine = V1LLMEngine.from_engine_args(engine_args=engine_args)
engine = LLMEngine.from_engine_args(engine_args=engine_args)
sampling_params = SamplingParams(max_tokens=1)
engine.add_request("0", "foo", sampling_params)