4 ms·

Clean implementations in TF2 of recent generative models – by Sarus Tech

• ngrislain 6y ago
  # Sarus published models Sarus implementation of classical ML models. The models are implemented using the Keras API of tensorflow 2. Vizualization are implemented and can be seen in tensorboard. The required packages are managed with pipenv and can be installed using pipenv install. Please see the pipenv documentation for more information. ## Philosophy These models' implementations are intended to be easy to read and to adapt by making use of the latest Tensorflow 2 library and Keras API. ## Basic usage To install and train a model. pipenv install pipenv shell python train.py To visualize losses and reconstructions. tensorboard --logdir ./logs/ ## Available models - Simple Autoencoder - Variational Autoencoder (VAE) - Vector Quantized Autoencoder (VQ-VAE) - PixelCNN - Gated PixelCNN - PixelCNN++ - Conditional Neural Processes - PixelSNAIL