Update backend/query_llm.py

backend/query_llm.py

@@ -50,7 +50,7 @@ def format_prompt(message: str, api_kind: str):
         raise ValueError("API is not supported")
 
 
-def generate_hf(prompt: str, history: str, temperature: float = 0.9, max_new_tokens: int = 256,
+def generate_hf(prompt: str, history: str, temperature: float = 0.9, max_new_tokens: int = 3000,
              top_p: float = 0.95, repetition_penalty: float = 1.0) -> Generator[str, None, str]:
     """
     Generate a sequence of tokens based on a given prompt and history using Mistral client.