Doubt about the patch_size parameter

[Aarsh2001]

I wanted to know if we are only training on one random patch of a whole image, rather than training on all the patches of the image. I could be wrong, can anyone help clarify this, thanks in advance 😁