fix ci

Signed-off-by: wangli <wangli858794774@gmail.com>

Author: Potabk

fusion_result.json

@@ -0,0 +1 @@
+null

tests/conftest.py

@@ -53,11 +53,7 @@ def cleanup_dist_env_and_memory():
     destroy_distributed_environment()
     gc.collect()
     torch.npu.empty_cache()
-    try:
-        torch._C._host_emptyCache()
-    except AttributeError:
-        logger.warning(
-            "torch._C._host_emptyCache() only available in Pytorch >=2.5")
+
 
 class VllmRunner:
 

tests/test_offline_inference.py

@@ -38,13 +38,7 @@
 @pytest.mark.parametrize("use_v1", ["1", "0"])
 @pytest.mark.parametrize("dtype", ["half", "float16"])
 @pytest.mark.parametrize("max_tokens", [5])
-def test_models(
-    model: str,
-    use_v1: str,
-    dtype: str,
-    max_tokens: int,
-    monkeypatch: pytest.MonkeyPatch
-) -> None:
+def test_models(model: str, use_v1: str, dtype: str, max_tokens: int, monkeypatch: pytest.MonkeyPatch) -> None:
     with monkeypatch.context() as m:
         m.setenv("VLLM_USE_V1", use_v1)
         if use_v1 == '1':
@@ -69,13 +63,9 @@ def test_models(
 @pytest.mark.parametrize("model, distributed_executor_backend", [
     ("Qwen/QwQ-32B", "mp"),
 ])
-def test_models_distributed(
-    vllm_runner,
-    use_v1: str,
-    model: str,
-    distributed_executor_backend: str,
-    monkeypatch: pytest.MonkeyPatch
-) -> None:
+def test_models_distributed(vllm_runner, use_v1: str, model: str,
+        distributed_executor_backend: str,
+        monkeypatch: pytest.MonkeyPatch) -> None:
     with monkeypatch.context() as m:
         m.setenv("VLLM_USE_V1", use_v1)
         if use_v1 == '1':