Use Keras Core with TensorFlow, PyTorch, and JAX backends
Introduction
Overview of Keras Core
Install the required dependencies
Run Keras Core
Next Steps
Use Keras Core with TensorFlow, PyTorch, and JAX backends
Who is this for?
This is an introductory topic for engineers who want to create a neural network model on Arm machines.
What will you learn?
Upon completion of this Learning Path, you will be able to:
- Create a simple neural network model using Keras Core.
- Train and evaluate your neural network model with different backends.
- Generate predictions with the trained model.
Prerequisites
Before starting, you will need the following:
- Basic Machine Learning knowledge
- An Arm based instance from a cloud service provider, an on-premises Arm server, or a Linux virtual machine on your Arm device
- Familiarity with SSH, the Linux command line, and basic system administration tasks
Summary
This summary was drafted with an approved AI-assisted workflow and reviewed by Arm contributors before publication. Human technical review remains part of the process so the final page reflects engineering rigor, accuracy, and Arm editorial standards.
You’ll build and run a compact neural network with Keras Core on an Arm-based Ubuntu server. First, you’ll prepare Python and define and execute a model that trains, evaluates, and predicts. Then, you’ll switch among the TensorFlow, PyTorch, and JAX backends. You’ll validate the complete workflow either locally or on an Arm-based instance over SSH by checking the printed training, evaluation, and prediction outputs.
Frequently asked questions
These FAQs were drafted with an approved AI-assisted workflow and reviewed by Arm contributors before publication. Human technical review remains part of the process so the final page reflects engineering rigor, accuracy, and Arm editorial standards.
After selecting a backend, start the run and check the initial console output to confirm that the chosen backend is in use. If it’s not the expected backend, repeat backend selection before training.
Create and activate the virtual environment that you prepared during dependency setup. Within the activated environment, save the script as
ml.py in a working directory. To run the script, run python ml.py.The run prints training progress and evaluation metrics, then shows predictions from the trained model. Seeing metrics and prediction values confirms that the end-to-end workflow completed.
Confirm that
keras_core is installed in your active Python environment and that the environment is activated. On Ubuntu 22.04, also verify that python3-pip and python3-venv are installed if you use the system Python.Use a Python version supported by the required dependencies. To stay consistent with the Learning Path, use Python 3.10 or 3.11. If you want to use a newer version such as Python 3.12, check package support because TensorFlow and PyTorch might not provide packages for it.