search

NVlabs / NVAE

The Official PyTorch Implementation of "NVAE: A Deep Hierarchical Variational Autoencoder" (NeurIPS 2020 spotlight paper)
https://arxiv.org/abs/2007.03898
Other
999 stars 163 forks source link

num_latent_scales on ImageNet 32x32 dataset #20

Closed liuyuyuil closed 3 years ago

liuyuyuil commented 3 years ago

Thanks for sharing your code. I have found that the hyper-parameter 'num_latent_scales' on ImageNet 32x32 and CIFAR-10 has been set to be 1, which is quite smaller than that of 5 on FFHQ and CelebA HQ dataset. What will happen if we set this number as 5 on ImageNet 32x32 dataset ? Thanks~

arash-vahdat commented 3 years ago

The number of latent scales represents the number of latent scales in the latent space. For CIFAR-10, we have only one scale at 16x16 dimensions. We didn't see improvements in terms of log-likelihood for using multi-scale latent space.