[Bugfix] Fix snapshot download in serving benchmark (vllm-project#6318)

Signed-off-by: LeiWang1999 <leiwang1999@outlook.com>

Author: ywang96

benchmarks/backend_request_func.py

@@ -390,17 +390,17 @@ def remove_prefix(text: str, prefix: str) -> str:
     return text
 
 
-def get_model(pretrained_model_name_or_path: str):
+def get_model(pretrained_model_name_or_path: str) -> str:
     if os.getenv('VLLM_USE_MODELSCOPE', 'False').lower() == 'true':
         from modelscope import snapshot_download
-    else:
-        from huggingface_hub import snapshot_download
-
-    model_path = snapshot_download(
-        model_id=pretrained_model_name_or_path,
-        local_files_only=huggingface_hub.constants.HF_HUB_OFFLINE,
-        ignore_file_pattern=[".*.pt", ".*.safetensors", ".*.bin"])
-    return model_path
+
+        model_path = snapshot_download(
+            model_id=pretrained_model_name_or_path,
+            local_files_only=huggingface_hub.constants.HF_HUB_OFFLINE,
+            ignore_file_pattern=[".*.pt", ".*.safetensors", ".*.bin"])
+
+        return model_path
+    return pretrained_model_name_or_path
 
 
 def get_tokenizer(