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

[BugFix] Fix async scheduling + chunked prefill + preemption (#28787)

Browse Source 
Signed-off-by: Nick Hill <nhill@redhat.com>
This commit is contained in:
Nick Hill
2025-11-16 14:46:46 -08:00
committed by GitHub
parent 03ee48111d
commit 80b6080ddc
3 changed files with 8 additions and 9 deletions
Download Patch File Download Diff File Expand all files Collapse all files

10
tests/v1/e2e/test_async_scheduling.py
View File

@@ -65,9 +65,8 @@ def test_without_spec_decoding(
(True, "mp", True, None, False),
(True, "uni", True, None, False),
(False, "mp", True, None, True),
# Async scheduling + preemption + chunked prefill needs to be fixed (WIP)
# (True, "mp", True, None, True),
# (True, "uni", True, None, True),
(True, "mp", True, None, True),
(True, "uni", True, None, True),
]
run_tests(
@@ -103,9 +102,8 @@ def test_with_spec_decoding(monkeypatch: pytest.MonkeyPatch):
(False, "mp", True, spec_config_short, True),
(True, "uni", True, spec_config, False),
(True, "uni", True, spec_config_short, False),
# Async scheduling + preemption + chunked prefill needs to be fixed (WIP)
# (True, "mp", True, spec_config, True),
# (True, "uni", True, spec_config_short, True),
(True, "mp", True, spec_config, True),
(True, "uni", True, spec_config_short, True),
]
run_tests(