Open caihaunqai opened 1 year ago
@caihaunqai Have you successfully generated images based on the Upsampling 256x256 model? I found that in this model, the out channel number is 6, why? I think that should be 3, like the RGB figure, hope to receive your response, thank you.
I get it, the first 3 channels are the mean of X{t-1}, the last 3 channels are the variance of X{t-1}
Thank you for your excellent job. I wonder which dataset is "Upsampling 256x256 model (280M parameters, trained for 500K iterations)" is trained on?