One of the key challenges in the fast moving artificial intelligence space is the speed of change in the underlying platforms, libraries and hardware.

Here at LogicalSpark we focus on using the right tool for job, so we have to keep up to date with changes and experiment constantly.

That’s why we often end of packing our own versions of tools with customisations just for us - or our clients. That’s why we’ve decided to start publishing some of these builds so you don’t have to.

With that in mind, we’re pleased to share custom wheels of TensorFlow with CUDA 9 and Cunn 7 support for Ubuntu 17.04 and 17.10 here.