This repository contains an approach to implement Grammar Restricted VAE For encoding and decoding arithmetic equations. @@ University of Stuttgart Deep Learning Lab

+ All the model classes are present in models.py
+ The training method for each model is present in train.py
+ Evaluation method for each model is present in eval.py
+ Visulization methods are present in visualize.py
+ Code is tightly integrasted with mlflow library which is used for experiment tracking and hyperparamter tuning.
+ List of hyperparamters are present in hyperparameters/params.yaml which can be modified as required.

• Main Notebook for Grammar VAE: Character&GrammarVAE.ipynb
• Notebook for Understanding AE and VAE Concepts using MNIST: UnderstandingAutoEncoders.ipynb