Open SA-j00u opened 4 months ago
Is it possible to train ESRGAN on non-square dataset without cliping? i have some 4x3 dataset that completely irrational to split into squares
Is it possible to train ESRGAN on non-square dataset without cliping? i have some 4x3 dataset that completely irrational to split into squares