A Differentiable Programming System to Bridge Machine Learning and Scientific Computing

TL;DR

This paper introduces Zygote, a differentiable programming system in Julia that enables gradient computation for complex programs, bridging machine learning and scientific computing with high performance and broad language support.

Contribution

The paper presents Zygote, a novel differentiable programming system capable of handling general program structures in Julia, facilitating integration of scientific computing and machine learning.

Findings

Supports almost all language constructs including control flow and recursion

Enables high-performance code compilation without user intervention

Facilitates incorporation of extensive libraries into differentiable programs

Abstract

Scientific computing is increasingly incorporating the advancements in machine learning and the ability to work with large amounts of data. At the same time, machine learning models are becoming increasingly sophisticated and exhibit many features often seen in scientific computing, stressing the capabilities of machine learning frameworks. Just as the disciplines of scientific computing and machine learning have shared common underlying infrastructure in the form of numerical linear algebra, we now have the opportunity to further share new computational infrastructure, and thus ideas, in the form of Differentiable Programming. We describe Zygote, a Differentiable Programming system that is able to take gradients of general program structures. We implement this system in the Julia programming language. Our system supports almost all language constructs (control flow, recursion, mutation, etc.) and compiles high-performance code without requiring any user intervention or refactoring to stage computations. This enables an expressive programming model for deep learning, but more importantly, it enables us to incorporate a large ecosystem of libraries in our models in a straightforward way. We discuss our approach to automatic differentiation, including its support for advanced techniques such as mixed-mode, complex and checkpointed differentiation, and present several examples of differentiating programs.