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, 14:45 UTC
[Bugfix] Fix test mocks after SM100 restriction in #38730 (#38791)
Signed-off-by: Stefano Castagnetta <scastagnetta@nvidia.com> Co-authored-by: Claude <noreply@anthropic.com>
Stefano Castagnetta committed
Apr 2, 2026 at 19:12 UTC
58262dec6e8190f321edd088e937d9b51f7ed064
1 file changed
+3
-3
tests/kernels/attention/test_use_trtllm_attention.py
+3
-3
@@ -62,7 +62,7 @@ def test_supports_batch_invariant_disables():
62
63
@patch("vllm.envs.VLLM_BATCH_INVARIANT", False)
64
@patch(
65
- "vllm.utils.flashinfer.current_platform.is_device_capability_family",
65
+ "vllm.utils.flashinfer.current_platform.is_device_capability",
66
return_value=True,
67
)
68
@patch("vllm.utils.flashinfer.has_nvidia_artifactory", return_value=True)
@@ -72,7 +72,7 @@ def test_supports_sm100_with_artifactory(_art, _cap):
72
73
@patch("vllm.envs.VLLM_BATCH_INVARIANT", False)
74
@patch(
75
- "vllm.utils.flashinfer.current_platform.is_device_capability_family",
75
+ "vllm.utils.flashinfer.current_platform.is_device_capability",
76
return_value=False,
77
)
78
def test_supports_non_sm100_platform(_cap):
@@ -81,7 +81,7 @@ def test_supports_non_sm100_platform(_cap):
81
82
@patch("vllm.envs.VLLM_BATCH_INVARIANT", False)
83
@patch(
84
- "vllm.utils.flashinfer.current_platform.is_device_capability_family",
84
+ "vllm.utils.flashinfer.current_platform.is_device_capability",
85
return_value=True,
86
)
87
@patch("vllm.utils.flashinfer.has_nvidia_artifactory", return_value=False)