[serving] fix stop token

imoneoi · imoneoi · commit edaa01111a27 · 2023-08-03T19:10:40.000+08:00
diff --git a/ochat/serving/openai_api_protocol.py b/ochat/serving/openai_api_protocol.py
@@ -59,7 +59,7 @@ class ChatCompletionRequest(BaseModel):
     top_p: Optional[float] = 1.0
     n: Optional[int] = 1
     max_tokens: Optional[int] = 768
-    stop: Optional[Union[str, List[str]]] = Field(default_factory=list)
+    stop: Optional[Union[str, List[str]]] = None
     stream: Optional[bool] = False
     presence_penalty: Optional[float] = 0.0
     frequency_penalty: Optional[float] = 0.0
diff --git a/ochat/serving/openai_api_server.py b/ochat/serving/openai_api_server.py
@@ -160,7 +160,7 @@ async def create_chat_completion(raw_request: Request):
             frequency_penalty=request.frequency_penalty,
             temperature=request.temperature,
             top_p=request.top_p,
-            stop=[model.eot_token],
+            stop=request.stop,
             max_tokens=request.max_tokens
         )
     except ValueError as e:

Original file line number	Diff line number	Diff line change
`@@ -160,7 +160,7 @@ async def create_chat_completion(raw_request: Request):`
`160`	`160`	`frequency_penalty=request.frequency_penalty,`
`161`	`161`	`temperature=request.temperature,`
`162`	`162`	`top_p=request.top_p,`
`163`		`- stop=[model.eot_token],`
	`163`	`+ stop=request.stop,`
`164`	`164`	`max_tokens=request.max_tokens`
`165`	`165`	`)`
`166`	`166`	`except ValueError as e:`