模型部署（Model deployment）

[Aanlifire]

Select Topic Area

Question

Body

有人部署过qwen2-vl模型吗，通过fastapi的方式接入到dify中测试的时候不会显示图片识别结果，但是使用curl和postman测试API的时候可以正常输出图片信息，有人遇到过这种问题吗，应该怎么解决呢？
Has anyone deployed the qwen2-vl model before? When testing it in dify through fastAPI, it does not display image recognition results. However, when testing the API with curl and postman, it can output image information normally. Has anyone encountered this problem before? How can we solve it?

[Aanlifire]

Hi, thank you very much for your suggestion, I have solved this problem