MUNIT paper AdaptiveInstanceNorm2d and LayerNorm issue

I have an equation about the decoder normalization type. in the follwwing repo

Is the normalization step done in the Residual Block parts of the decoder or in the up-sampling part of the decoder, I misunderstand the difference between the AdaptiveInstanceNorm2d and the LayerNorm in the code since the paper said use the AdaIN not mention the LayerNorm ?
in LayerNorm where are the gamma and beta come from? Is it from MLP or what?

anyone can give me a hint, thank you in advance