https://github.com/DingKe/nn_playground

Experimental keras implementation of novel neural network structures
https://github.com/DingKe/nn_playground

deep-learning deep-neural-networks keras neural-networks

Last synced: 8 days ago
JSON representation

Experimental keras implementation of novel neural network structures

• Host: GitHub
• URL: https://github.com/DingKe/nn_playground
• Owner: DingKe
• License: mit
• Created: 2016-12-28T06:49:00.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-09-29T01:28:30.000Z (over 5 years ago)
• Last Synced: 2024-02-28T21:33:00.702Z (4 months ago)
• Topics: deep-learning, deep-neural-networks, keras, neural-networks
• Language: Python
• Homepage:
• Size: 2.86 MB
• Stars: 430
• Watchers: 25
• Forks: 153
• Open Issues: 11
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Lists

• awesome-deeplearning-resources - Experimental implementation of novel neural network structures

README

        
# binarynet

Binary Networks

# xnornet

XNOR Networks

# ternarynet

Ternary Networks

# qrnn

Quasi-Recurrent Nueral Networks

# vae

Variational Auto-Encoder

# gcnn

Gated Convolutional Nueral Networks

# weighnorm

Weight Normalization

# layernorm

Layer Normalization

# wgan 

Wasserstein GAN

# lsgan 

Least Squares GAN

# glsgan

Generalized Loss Sensitive GAN

# focal_loss

Focal Loss

# senet

Squeeze-and-Excitation Networks

Note:

By default, all keras scripts run on tensorflow backend with image_data_format 'channels_first'.