qwen2.5模型时，调用agent不返回结果

[zzlalb]

问题描述 / Problem Description
使用qwen2.5模型时，调用agent不返回结果

复现问题的步骤 / Steps to Reproduce

1. 启用数学计算器agent
2. 显示agent过程
3. 输入“计算8500*8500”

预期的结果 / Expected Result
返回agent的结果

实际结果 / Actual Result
模型仅输出：“Thought: 这是一个简单的乘法计算，可以使用calculate API来执行。 Action: calculate Action Input: {"expression": "8500 * 8500"}”

环境信息 / Environment Information

• Langchain-Chatchat 版本号: 0.3.1.2
• 部署方式：dev deployment, pip install
• 使用的模型推理框架）： Ollama
• 使用的 LLM 模型： Qwen2.5
• 使用的 Embedding 模型：bge-m3
• 操作系统及版本 / Operating system and version: Windows
• Python 版本 / Python version: 3.10
• 推理使用的硬件（GPU / CPU / MPS / NPU 等） / Inference hardware (GPU / CPU / MPS / NPU, etc.): GPU

附加信息 / Additional Information
截图：

[dosubot]

Hello @zzlalb! 👋 I'm Dosu, your friendly bot here to help with bugs, answer questions, and assist you in becoming a contributor. I'm here while you wait for a human maintainer. Let’s tackle this together! 😊

The issue you're experiencing with the Qwen2.5 model not returning results when using the agent for calculations might be related to known issues with agent interactions in Langchain-Chatchat. In version 0.3.1.2, there have been improvements in agent functionality, but some problems might still persist, especially with tool calls during streaming ^[1][2].

Here are a few steps you can take to troubleshoot and potentially resolve the issue:

1. Check Tool Integration: Ensure that the 'calculate API' is properly integrated and accessible by the agent. The agent should be able to call this tool to perform calculations ^[3].

2. Review Logs: Use the logging and debugging features to trace the agent's execution. This can help identify if there are any errors or issues during the tool call ^[4].

3. Update or Reconfigure: Make sure that your setup is correctly configured and consider updating any dependencies or configurations that might affect the agent's ability to execute the tool.

4. Test with Different Configurations: If possible, test the agent with different configurations or environments to see if the issue persists. This can help determine if the problem is specific to your current setup.

If these steps do not resolve the issue, it might be beneficial to look into the LangGraph-Chatchat implementation, which addresses some of the agent-related issues ^[5].

To continue talking to Dosu, mention @dosu.

---

^{Help Dosu learn! Give it feedback: Great Response | Irrelevant Answer | Incorrect Sources | Too Verbose | Hallucination | Bug Report | Other}