Hi Sergei, thanks for your contribution! I am interested in this fast face generation. I wonder if you have ever benchmarked the performance of this model, have the latency on resolution 1024 or 512?
Also What is the different between this one and your another contribution MobileStyleGAN?