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] Remove faulty test for oot attention backend (#22286)
Signed-off-by: mgoin <mgoin64@gmail.com>
Michael Goin committed
Aug 5, 2025 at 20:05 UTC
35509fc5be5d840e84717ff24bba6bdd5cc33d77
1 file changed
-10
tests/plugins_tests/test_platform_plugins.py
-10
@@ -4,9 +4,7 @@
4
import pytest
5
import torch
6
7
-from vllm.attention.selector import get_attn_backend
7
from vllm.plugins import load_general_plugins
9
-from vllm.utils import STR_BACKEND_ENV_VAR, STR_INVALID_VAL
8
9
10
def test_platform_plugins():
@@ -27,14 +25,6 @@ def test_platform_plugins():
25
f" is loaded. The first import:\n{_init_trace}")
26
27
30
-def test_oot_attention_backend(monkeypatch: pytest.MonkeyPatch):
31
- # ignore the backend env variable if it is set
32
- with monkeypatch.context() as m:
33
- m.setenv(STR_BACKEND_ENV_VAR, STR_INVALID_VAL)
34
- backend = get_attn_backend(16, torch.float16, "auto", 16, False)
35
- assert backend.get_name() == "Dummy_Backend"
36
-
37
-
28
def test_oot_custom_op(monkeypatch: pytest.MonkeyPatch):
29
# simulate workload by running an example
30
load_general_plugins()