Abstract: 

Why should we try to unify the ML frameworks? Won't we just create a new incompatible standard and make the ML fragmentation even worse? I will argue that the answer to these sensible and important questions is no.

Session Outline:

Ivy as a Transpiler
A hands-on tutorial, showing how Ivy can be used to transpile any ML code to run in any other ML framework, with the addition of a single function decorator. We will use DeepMind's PerceiverIO implementation as an example.

Ivy as a Framework
A hands-on tutorial, showing how models can be implemented in Ivy directly, which can then be run with any ML framework in the backend, without the need to transpile any code.

Round Up
Concluding remarks on the future directions of Ivy, and what our immediate Roadmap looks like.

Background Knowledge:

ML Frameworks (JAX, TensorFlow, PyTorch)
Compilers and Intermediate Representations
Python3

Bio: 

Daniel Lenton is the creator of Ivy, which is an open-source framework with an ambitious mission to unify all other ML frameworks. Prior to starting Ivy, Daniel was a PhD student at Imperial College London, where he published research in the areas of machine learning, robotics and computer vision.