Skip to content

Latest commit

 

History

History
35 lines (19 loc) · 1.01 KB

README.md

File metadata and controls

35 lines (19 loc) · 1.01 KB

Pytorch Implementation of Glow and conditonal Glow and PixelCNN. Based on the paper:

Density estimation using Real NVP
Diederik P. Kingma, Prafulla Dhariwal
arXiv:1807.03039

Conditional Image Generation with PixelCNN Decoders
Aaron van den Oord, Nal Kalchbrenner, Oriol Vinyals, Lasse Espeholt, Alex Graves, Koray Kavukcuoglu
arXiv:1606.05328

The results for conditional Glow samples can be found in the gen_imgs folder.

MNIST

Epoch 1

Samples at Epoch 1000

Epoch 10

Samples at Epoch 10000

Epoch 20

Samples at Epoch 20000

Epoch 50

Samples at Epoch 50000

Usage

The entire code is self contained in the Jupyter notebook,just run the cells sequentially. It is made this way for ease of training on Google Colab.