A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[Bugfix][CI] Fix `Distributed Tests (4 GPUs)` async_sched+ray test (#27195)
Signed-off-by: NickLucche <nlucches@redhat.com>
Nicolò Lucchesi committed
Oct 20, 2025 at 18:34 UTC
f9e7ad54000fb982248564591007e038304f190e
1 file changed
+3
tests/v1/distributed/test_async_llm_dp.py
+3
@@ -78,6 +78,9 @@ async def generate(
78
async def test_load(
79
output_kind: RequestOutputKind, data_parallel_backend: str, async_scheduling: bool
80
):
81
+ if async_scheduling and data_parallel_backend == "ray":
82
+ # TODO(NickLucche) Re-enable when async scheduling is supported
83
+ pytest.skip("Async scheduling is not supported with ray")
84
stats_loggers = {}
85
86
@dataclass