Update app.py

app.py

@@ -72,6 +72,7 @@ def chat_completions():
     def generate():
         nonlocal messages
         full_response = ""
+        final_finish_reason = None
         while True:
             conversation = "\n".join([f"{msg['role']}: {msg['content']}" for msg in messages])
             conversation += "\n请关注并回复user最近的消息并避免总结对话历史的回答"
@@ -98,15 +99,19 @@ def chat_completions():
                     final_data = json.loads(event.data)
                     finish_reason = final_data.get('responseMessage', {}).get('finish_reason', 'stop')
                     if finish_reason == 'length':
-                        # 如果因为长度被截断，添加已生成的回复到消息列表，并继续生成
                         messages.append({"role": "assistant", "content": full_response})
                         messages.append({"role": "user", "content": "请继续你的输出，不要重复之前的内容"})
                         break
                     else:
+                        final_finish_reason = finish_reason
                         yield f"data: {json.dumps(format_openai_response('', finish_reason))}\n\n"
                         yield "data: [DONE]\n\n"
                         return
 
+        if final_finish_reason is None:
+            yield f"data: {json.dumps(format_openai_response('', 'stop'))}\n\n"
+            yield "data: [DONE]\n\n"
+
     if stream:
         return Response(stream_with_context(generate()), content_type='text/event-stream')
     else: