Warning: This post contains spoilers for the first episode of The White Lotus Season 3. Who would’ve thought that the ...

We describe a new training methodology for generative adversarial networks. The key idea is to grow both the generator and discriminator progressively: starting from a low resolution, we add new ...