Installation Instructions
The Extension for Scikit-learn* can be easily installed as a Python package under name scikit-learn-intelex from the most
common distribution channels for Python. For GPU support, an additional package scikit-learn-intelex-gpu is
also made available through the same channels - see GPU support for more information.
Binary wheels and conda packages are provided for the x86-64 architecture, for both Windows* and Linux*, for the Python versions supported by current versions of scikit-learn - see About the Extension for Scikit-learn* for more information.
Installing from PyPI
To install with pip from the Python Package Index (PyPI), execute the following command on a terminal:
pip install scikit-learn-intelex
Tip
To prevent version conflicts, we recommend creating and activating a new virtual environment for Extension for Scikit-learn*.
Tip
Wheels are also available through Intel’s index: https://software.repos.intel.com/python/pypi
Installing from conda-forge
To install the Extension for Scikit-learn* in a fresh conda environment, execute the following command in a terminal:
conda create -n sklex -c conda-forge --override-channels scikit-learn-intelex
conda activate sklex
To install in an existing environment:
conda install -c conda-forge --override-channels scikit-learn-intelex
Hint
It is advisable to use the miniforge installer for conda/mamba, as it comes with
conda-forge as the default channel.
Warning
The main Anaconda channel also provides distributions of scikit-learn-intelex, but it does not provide the latest versions, nor does
it provide GPU-enabled builds. It is highly recommended to install scikit-learn-intelex from either conda-forge or from Intel’s channel instead.
Tip
The Extension for Scikit-learn* is also available at Intel’s conda channel: https://software.repos.intel.com/python/conda
Packages from the Intel channel, which include also optimized versions of NumPy and SciPy, are meant to be used together with
dependencies from the conda-forge channel, and might not work correctly when used in an environment where packages from the
anaconda default channel have been installed.
Building from Sources
The Extension for Scikit-learn* is a fully open-source package (link to source code) which can be compiled from source with standard compiler toolkits.
See Building from Source for details.