Skip to content

AXLearn

Scale deep learning with a JAX-based library

Visit official site

What AXLearn is

The library is described as an Apple deep learning framework on JAX and XLA that handles large scale training through reusable configs.

What you can do with it

  • Compose models from shared building blocks
  • Describe work on a global virtual machine
  • Scale training across many accelerators

Application scenarios

  • Train a language model from a Transformer
  • Run image classification on many GPUs
  • Join components into a speech pipeline

Pros & cons

✓ What we like

  • Public cloud deployment is supported
  • Vision and speech tasks are covered

! What to watch out for

  • Models are composed from config building blocks
  • Training targets large accelerator clusters

FAQ

What is it built on?

The material names JAX and XLA.

Which fields are listed?

Natural language processing, computer vision and speech recognition are named.

Last reviewed: 2026-11-15

More AI coding tools tools

View all →

How we review