A look at the CPython execution model and showing two ways to run CUDA from Python, using ctypes for raw access and PyTorch custom operators for deeper integration.

Thrilled to announce our team’s victory at this weekend’s e/acc London hackathon!

We are looking for people to take part in a closed pilot of a new social platform we’re building – Futon. We think that 1) meeting new people sucks, 2) planning in advance isn’t fun, and 3) we all spend far too much time online.

It might be confusing at first how to reconcile the principles of Bayesian Inference with the framework of neural networks. Especially given the size of modern architectures and the nature of stochastic gradient based optimization which are usually not covered in Bayesian statistics resources.

Our paper ‘Hierarchical VampPrior Variational Fair Auto-Encoder’ got accepted to the ICML Theoretical Foundations and Applications of Deep Generative Models workshop. This paper is part of my thesis on Fair Deep Latent Variable Models under the supervision of Dr. Jakub Tomczak.

In most classic Machine Learning problems we are interested in learning a mapping from the input data to a label, more recently however, a lot of interest has sparked in the field of generative modelling. We will look at one of the most popular models in depth, the Variational Autoencoder.