fix

MeetKai · Aug 31, 2024 · 441506a · 441506a
1 parent 856a764
commit 441506a
Showing 1 changed file with 3 additions and 5 deletions.
diff --git a/server_vllm.py b/server_vllm.py
@@ -138,13 +138,11 @@ async def create_chat_completion(raw_request: Request):
 
     logger.info(f"args: {args}")
 
-    if args.served_model_name is not None:
-        logger.info(
-            "args.served_model_name is not used in this service and will be ignored. Served model will consist of args.model only."
-        )
-
     served_model = [args.model]
 
+    if args.served_model_name is not None:
+        served_model += args.served_model_name
+
     engine_args = AsyncEngineArgs.from_cli_args(args)
     # A separate tokenizer to map token IDs to strings.
     tokenizer = get_tokenizer(