Hello! Thank you very much for this, it seems really good!
I am wondering how the style adapter trained, for example, I have van Gogh images， then the input of adapter is the van Gogh, but what is the loss( is the target generation to reconstruct the van Gogh image?)
thanks for answering
quanh1990
commented
1 year ago
Hello! Thank you very much for this, it seems really good! I am wondering how the style adapter trained, for example, I have van Gogh images， then the input of adapter is the van Gogh, but what is the loss( is the target generation to reconstruct the van Gogh image?) thanks for answering