liveBook · Manning

Keras 3

As explained in Deep Learning with JAX

A multi-backend neural-network library that can run with JAX, TensorFlow, and PyTorch backends.

Need one neural-network workflow that can move between JAX, TensorFlow, and PyTorch? Keras 3 provides a multi-backend interface for running Keras workflows on any of these systems. section 12.1.1 section 2.8

Beyond standalone model building, Keras components such as layers, models, and metrics can integrate with lower-level JAX, TensorFlow, and PyTorch workflows. This makes Keras 3 a portability layer alongside more backend-specific libraries such as Flax and Equinox. section 12.1.1

Keras 3 is useful when you want to:

  • Run Keras workflows with JAX, TensorFlow, or PyTorch as the backend. section 12.1.1
  • Add Keras layers, models, or metrics to an existing low-level framework workflow. section 12.1.1
  • Compare a unified interface with framework-specific choices such as Flax or Equinox. section 12.1.1

How Deep Learning with JAX covers it

How deep it goes
Defines backend portability and component integration but does not detail implementation internals or training procedures.
Examples
Uses layers, models, and metrics as portable components within lower-level framework workflows.
What you can do after
Evaluate/understand where a multi-backend Keras interface fits among JAX, TensorFlow, and PyTorch libraries.
How it connects
Connects neural-network libraries with JAX, TensorFlow, PyTorch, Flax, Equinox, and Ivy.
Cover of Deep Learning with JAX
liveBookmanning.com
Read Deep Learning with JAX on liveBook Get the book at manning.com

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Deep Learning with JAX. The section links point to the parts of the book it draws on.