Implementing Variational Autoencoders

Collecting links

• Variational autoencoder implementation in keras with a custom training step
• Custom Training Using the Strategy Class.
• Withing a custom loss function, it's sometimes necessary to get the epoch number.
• Some reserch in determining, controlling the VAE loss function to avoid posterior collapse in Re-balancing Variational Autoencoder Loss for Molecule Sequence Generation.
• A review of Dropout as applied to RNNs
• Dropout with LSTM Networks for Time Series Forecasting