In the example provided in the code, the image generation function only supports text input. Can the image generation support multimodal input? How should image input be handled?

[ShenXiongquan]

In the example provided in the code, the image generation function only supports text input. Can the image generation support multimodal input? How should image input be handled?

[ShenXiongquan]

在代码给的例子中，生成图像的功能中只支持文本输入，请问生成图像可以支持多模态的输入吗，如果可以，应该如何处理图像输入呢？