An Implementation of the paper "Variational Deep Embedding: An Unsupervised and Generative Approach to Clustering" (https://arxiv.org/abs/1611.05148), with additional code for a Variational Deep embedding model that incorperates triplet loss to utilise weak supervision on the clustering task. The code is writtent in Pytorch, with Pytorch-lightning modules for organaizing the code. There is also code for running experiments on the models with config.yaml files, and for running hyperparameter swipes, using Weights and Biases platform.