What is the current best GAN architecture for generating training images for DETECTION that has a publicly available code?

Question

I am currently looking for the best GAN architecture in which there is a publicly available Pytorch code. My questions would be:

• Anyone who has implemented G