MLX backend implementation for Keras

Local development

For development, you need the keras repository and the keras-mlx repository checked out locally. That's because the unit tests code is in the keras repository.

We first check out the main keras repository and the pluggable_backend branch.

gh repo clone keras-team/keras
cd keras
git checkout pluggable_backend
pip install -r requirements-common.txt
cd ..

Assuming you have a fork of keras-mlx, you will run the following. This also installs keras-mlx locally so that keras can find and import the keras-mlx module.

gh repo clone <your_github_handle>/keras-mlx
cd keras-mlx
pip install -r requirements.txt
pip install -e .
cd ..

Running tests happens from the root of the keras repository.

cd keras
KERAS_BACKEND=mlx pytest keras --ignore=keras/src/applications \
    --ignore=keras/src/wrappers

Formatting the code

The code is formatted and linted with Ruff, run through pre-commit. The first time you are setting up the repo, please run pre-commit install. Note that this needs to be done only once at the beginning.

Now, whenever you run git commit -m "<message>", the code is automatically formatted and linted. If there's any error, the commit will not go through. Please fix the error (most of the times, the error is fixed automatically by the formatter/linter) and re-run the following:

git add .
git commit -m "<message>"

In case you want to run the above manually on all files, you can do the following:

pre-commit run --all-files