Skip to content

Latest commit

 

History

History
executable file
·
27 lines (17 loc) · 731 Bytes

README.md

File metadata and controls

executable file
·
27 lines (17 loc) · 731 Bytes

PixelVAE

Code for the models in PixelVAE: A Latent Variable Model for Natural Images

Binarized MNIST

To train:

python models/mnist_pixelvae_train.py -L 12 -fs 5 -algo cond_z_bias -dpx 16 -ldim 16

To evaluate, take the weights of the model with best validation score from the above training procedure and then run

python models/mnist_pixelvae_evaluate.py -L 12 -fs 5 -algo cond_z_bias -dpx 16 -ldim 16 -w path/to/weights.pkl

Real-valued MNIST, LSUN Bedrooms, 64x64 ImageNet

To train, evaluate, and generate samples:

python pixelvae.py

By default, this runs on real-valued MNIST. You can specify different datasets or model settings within pixelvae.py.