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 5, 02:15 UTC
[CI Bug] fix flaky test (#39938)
Signed-off-by: yewentao256 <zhyanwentao@126.com>
Wentao Ye committed
Apr 15, 2026 at 17:20 UTC
7c636432c62e242d36bff056d5258474498d3a4e
1 file changed
+1
-1
tests/v1/kv_connector/unit/test_nixl_connector_hma.py
+1
-1
@@ -195,7 +195,7 @@ def test_fewer_blocks_with_hma(monkeypatch, model_name, sw_size):
195
llm_kwargs = {
196
"model": model_name,
197
"enforce_eager": True,
198
- "gpu_memory_utilization": 0.5,
198
+ "gpu_memory_utilization": 0.47,
199
"kv_transfer_config": kv_transfer_config,
200
"max_model_len": 2048,
201
# NOTE: Make sure HMA is enabled