llama3.2_11b_vision用MathVista_MINI测评时，由于不设置.env文件，会报错

[GoogleAlphaZero]

llama3.2_11b_vision用MathVista_MINI测评时，由于不设置.env文件，会报错。然后我又建立了.env文件，内部是空的，再执行，也一样报错

[GoogleAlphaZero]

MathVision_MINI也是一样，.env文件内的Openai 密钥为空，会报错

[GoogleAlphaZero]

MathVista和MathVision，必须要openai的密钥吗？还有哪几个模型必须要密钥，否则无法run通？

[isruihu]

跟模型没关系，跟数据集有关系

[PhoenixZ810]

Hello,

Thank you for reaching out.

I hope this message finds you well. I would like to inform you that certain complex benchmarks, such as MMVet and MathVista, require an OpenAI API key to retrieve accurate results. Should you wish to run these specific benchmarks, please ensure you have provided the necessary OpenAI API key. You can verify which benchmarks necessitate an OpenAI API key by reviewing the run.py script in our repository.

[GoogleAlphaZero]

好的，感谢