All Questions
1 question
0
votes
1
answer
833
views
Why does the latent space in Stable Diffusion have a shape of 64x64x3?
Since the encoding is performed by a Variational Autoencoder, the VAE encoder must output some mean and log variance that we can ...