Update app.py

app.py

@@ -34,6 +34,12 @@ class ChatCompletionRequest(BaseModel):
     messages: List[ChatMessage]
     temperature: float = 0.7
     stream: bool = False
+    frequency_penalty: float = 0.0
+    presence_penalty: float = 0.0
+    max_tokens: int = 512
+    seed: int = None
+    top_p: float = 1.0
+    tools: List[Any] = None
 
 def count_tokens(text: str) -> int:
     # Используем тот же алгоритм, что и в прямом API
@@ -116,17 +122,43 @@ async def get_models():
     except requests.RequestException as e:
         raise HTTPException(status_code=500, detail=str(e))
 
+def extract_system_prompt(messages: List[ChatMessage]) -> str:
+    """Извлекает system prompt из сообщений"""
+    for msg in messages:
+        if msg.role == "system":
+            return msg.content
+    return ""
+
+def get_conversation_history(messages: List[ChatMessage]) -> List[Dict[str, str]]:
+    """Преобразует сообщения в формат для Flowise"""
+    history = []
+    for msg in messages:
+        if msg.role in ["user", "assistant"]:
+            history.append({
+                "role": msg.role,
+                "content": msg.content
+            })
+    return history
+
 @app.post("/v1/chat/completions")
 async def create_chat_completion(request: ChatCompletionRequest):
     try:
-        # Получаем последнее сообщение из диалога
-        last_message = request.messages[-1]
-        if last_message.role != "user":
-            raise HTTPException(status_code=400, detail="Last message must be from user")
+        # Извлекаем system prompt
+        system_prompt = extract_system_prompt(request.messages)
+        
+        # Получаем историю диалога
+        conversation_history = get_conversation_history(request.messages)
+        
+        # Получаем последнее сообщение пользователя
+        last_user_message = next((msg for msg in reversed(request.messages) if msg.role == "user"), None)
+        if not last_user_message:
+            raise HTTPException(status_code=400, detail="No user message found")
         
         # Формируем запрос к Flowise
         flowise_request = {
-            "question": last_message.content
+            "question": last_user_message.content,
+            "system_prompt": system_prompt,
+            "conversation_history": conversation_history
         }
         
         # Засекаем время начала запроса
@@ -136,7 +168,7 @@ async def create_chat_completion(request: ChatCompletionRequest):
         response = requests.post(
             f"{FLOWISE_API_BASE_URL}/prediction/{FLOWISE_CHATFLOW_ID}",
             json=flowise_request,
-            timeout=10  # Уменьшаем таймаут до 10 секунд
+            timeout=10
         )
         response.raise_for_status()
         
@@ -152,7 +184,7 @@ async def create_chat_completion(request: ChatCompletionRequest):
             )
         
         # Подсчитываем токены
-        prompt_tokens = count_tokens(last_message.content)
+        prompt_tokens = count_tokens(last_user_message.content)
         completion_tokens = count_tokens(assistant_response)
         
         response = JSONResponse({