How to run OpenVINO 2022.3 LTS on Apple M1?

Raymond Lo, PhD
2 min readDec 22, 2022

--

It’s now part of the PyPI.

Here is what I did to get it running :). I hand (manually) installed the following packages one by one. And then run OpenVINO Notebooks as usual.

https://github.com/openvinotoolkit/openvino_notebooks

#install all packages
python3 -venv openvino2023
source openvino2023/bin/activate
pip install --upgrade pip
pip install openvino-dev==2022.3
pip install tensorflow-macos
pip install torch
pip install onnx
pip install torchvision
pip install jupyterlab
It worked :)

And here is the quick instruction on how I started the notebooks. Nothing unusual.

git clone https://github.com/openvinotoolkit/openvino_notebooks.git
cd openvino_notebooks
jupyter lab notebooks

And it works with webcam and all too! :).

I didn’t install NNCF and some of the deeper libraries, but that’s something maybe worth a try later on.

One of the open debates now is if running this with Rosetta 2 is going to be faster. :) It’s possible! and let’s try and see next time.

--

--

Raymond Lo, PhD
Raymond Lo, PhD

Written by Raymond Lo, PhD

@Intel - OpenVINO AI Software Evangelist. ex-Google, ex-Samsung, and ex-Meta (Augmented Reality) executive. Ph.D. in Computer Engineer — U of T.

No responses yet