diff --git a/.github/workflows/vllm_ascend_test.yaml b/.github/workflows/vllm_ascend_test.yaml
index 52ef401..7f8ac0e 100644
--- a/.github/workflows/vllm_ascend_test.yaml
+++ b/.github/workflows/vllm_ascend_test.yaml
@@ -115,7 +115,7 @@ jobs:
             pytest -sv tests/ops
             pytest -sv tests/compile
           else
-            pytest -sv tests/multicard/test_offline_inference_distributed.py
+            pytest -sv -k "QwQ" tests/multicard/test_offline_inference_distributed.py
             pytest -sv tests/ops
             pytest -sv tests/compile
           fi
@@ -128,7 +128,8 @@ jobs:
             pytest -sv tests/singlecard/test_offline_inference.py
             pytest -sv tests/ops
           else
-            pytest -sv tests/multicard/test_offline_inference_distributed.py
+            pytest -sv -k "QwQ" tests/multicard/test_offline_inference_distributed.py
+            pytest -sv -k "DeepSeek" tests/multicard/test_offline_inference_distributed.py
             pytest -sv tests/ops
           fi
 
diff --git a/tests/multicard/test_offline_inference_distributed.py b/tests/multicard/test_offline_inference_distributed.py
index dfc6675..26a3de5 100644
--- a/tests/multicard/test_offline_inference_distributed.py
+++ b/tests/multicard/test_offline_inference_distributed.py
@@ -28,10 +28,12 @@ import vllm  # noqa: F401
 from tests.conftest import VllmRunner
 
 os.environ["PYTORCH_NPU_ALLOC_CONF"] = "max_split_size_mb:256"
+os.environ["VLLM_USE_MODELSCOPE"] = "True"
 
 
 @pytest.mark.parametrize("model, distributed_executor_backend", [
     ("Qwen/QwQ-32B", "mp"),
+    ("deepseek-ai/DeepSeek-V2-Lite", "mp"),
 ])
 def test_models_distributed(model: str,
                             distributed_executor_backend: str) -> None: