fix

evaluator.py

@@ -167,8 +167,7 @@ In addition, the following function calls are also part of the reasoning steps.
 {"\n".join([step.to_prompt(ignore_output=True) for step in output_answer.function_calls])}
 ```
 
-Evaluation Rules:
-"""
+Evaluation Rules:\n"""
         for item in reasoning_items:
             prompt += f"{item.to_prompt()}\n"
         prompt += f"Now evaluate the reasoning steps based on the evaluation criteria, and give the score for each item in the range of 0 to the point the criteria worth."