diff --git a/configs/model_config.py.example b/configs/model_config.py.example
index 5b2574e91..308fb8ee5 100644
--- a/configs/model_config.py.example
+++ b/configs/model_config.py.example
@@ -70,7 +70,6 @@ llm_model_dict = {
     # 4.0 seconds as it raised APIConnectionError: Error communicating with OpenAI.
     # 需要添加代理访问(正常开的代理软件可能会拦截不上)需要设置配置openai_proxy 或者 使用环境遍历OPENAI_PROXY 进行设置
     "gpt-3.5-turbo": {
-        "local_model_path": "gpt-3.5-turbo",
         "api_base_url": "https://api.openai.com/v1",
         "api_key": os.environ.get("OPENAI_API_KEY"),
         "openai_proxy": os.environ.get("OPENAI_PROXY")
diff --git a/startup.py b/startup.py
index c8706e281..2ca62d932 100644
--- a/startup.py
+++ b/startup.py
@@ -404,14 +404,16 @@ def dump_server_info(after_start=False):
         processes["openai_api"] = process
 
     if args.model_worker:
-        process = Process(
-            target=run_model_worker,
-            name=f"model_worker({os.getpid()})",
-            args=(args.model_name, args.controller_address, queue, len(processes) + 1),
-            daemon=True,
-        )
-        process.start()
-        processes["model_worker"] = process
+        model_path = llm_model_dict[args.model_name].get("local_model_path", "")
+        if os.path.isdir(model_path):
+            process = Process(
+                target=run_model_worker,
+                name=f"model_worker({os.getpid()})",
+                args=(args.model_name, args.controller_address, queue, len(processes) + 1),
+                daemon=True,
+            )
+            process.start()
+            processes["model_worker"] = process
 
     if args.api:
         process = Process(