Allen

@happypepper

@happypepper Hi, thank you for your reply. I solve 2 streets instead of using a river network. And I calculated my exploitability of a turn case, the exploitability is around...

@happypepper I use k-means on the river round, and EMD on other rounds. I used the same bucketing in the reference papers. I noticed that in your code, you made...

@happypepper Hi, I just sent an email to you and described the method of generating river clusters.

> [https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/function_calling.py](https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/function_calling.py#L48) 请问这个文件里function call的prompt为什么会包含"result"和"return"呢，我理解assistant应该只会输出function name和arguments？

> > > [https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/function_calling.py](https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/function_calling.py#L36) > > > > > > 请问这个文件里function call的prompt为什么会包含"result"和"return"呢，我理解assistant应该只会输出function name和arguments？ > > result和return中间的内容是工具返回的结果。prompt里包含工具结果的目的是：让模型根据这次的工具结果判断接下来是要 总结工具结果就结束、还是继续调用其他工具。 嗯嗯，谢谢，不过我在使用这个prompt直接call模型(Qwen2-72B-Instruct-GPTQ-Int4)的时候，发现模型会直接想象一个结果填写到result后面，可能是我打开方式不对？我是把如下prompt format之后拼在system_prompt里，与[qwen_agent/llm/function_calling.py](https://github.com/QwenLM/Qwen-Agent/blob/main/qwen_agent/llm/function_calling.py#L154)中的拼接方式相同。 ``` FN_CALL_TEMPLATE_FMT_EN = """## When you need to call a tool,...