Skip to content

Latest commit

 

History

History
18 lines (13 loc) · 461 Bytes

README.md

File metadata and controls

18 lines (13 loc) · 461 Bytes

Experimental implementations of neural networks using causal-convolutions. All the neural network is built upon Gluon API of Apache MXNet.

Implementations

  1. Gated PixelCNN
  2. Conditional PixelCNN
  3. Dilated Convolution

ToDo

  • Caching hidden layers for fast generation
  • Logistic Mixture
  • VQ-VAE
  • WaveNet

References

  • A. van den Oord, et.al. "Conditional Image Generation with PixelCNN Decoders" NIPS2016