andy-96 / tensorflow_macos

TensorFlow for macOS 11.0+ accelerated using Apple's ML Compute framework.

Installing Tensorflow for M1

Preparation

1. Use Apple's Python 3.8.2 version. Verify this by checking which python3, it should be /usr/bin/python3
2. Deactivate anaconda environment, if it is activated
3. Download .tar.gz archive: https://github.com/apple/tensorflow_macos/releases (Current version: v0.1alpha1)
4. Change all files under tensorflow_macos/arm64 from *-cp38-macosx_11_0_*.whl to *-cp38-macosx_10_16_*.whl
5. Replace this repo's install script ./install_venv.sh with original one

Using Python's venv

1. Cd into your project repo
2. Create virtual environment: python3 -m venv .venv
3. Make sure the used python version in your venv is python 3.8 if you have other Python versions installed
4. Run <path-to-tensorflow_macos>/tensorflow_macos/install_venv.sh and use <path-to-project>/.venv as the venv folder
5. Activate venv: .venv/bin/activate

Using Pipenv

1. Cd into your project repo
2. Create virtual environment: pipenv --python 3.8
3. Run pipenv shell and copy the virtualenv path
4. Run ``<path-to-tensorflow_macos>/tensorflow_macos/install_venv.sh` and use the virtualenv path

Test whether it worked

1. Activate virtualenv
2. Run python
3. import tensorflow as tf
4. print(tf.constant(4))