How to setup Mac M2 Pro for Deep Learning in 2024

I recently migrated to a Mac M2 Pro from a Windows PC and was having a tough time of figuring out the Mac UI let alone how to use a Mac for Deep Learning. I think its safe to say at first I missed my PC and was seriously questioning my decision to switch to a mac book but I didn’t let myself lose sight of the fact that the reason I switched to a Mac was to utilise its top notch GPU for Deep Learning.

So, here’s all the steps you need to follow to have your PC set-up for DeepLearning, by installing pandas, numpy, matplotlib, scikit-learn, Tensorflow and Jupyter:

1. Install Homebrew
2. Install Miniconda
3. Create environment
4. Install Tensorflow and its dependencies
5. Confirm Tensorflow is accessing GPU

1) Installing Homebrew

Homebrew is a package manager, which doesn’t come pre-installed in Mac, so we install by running this command:

/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"

This comand can be found on https://brew.sh/.

2) Download & Install Miniconda

https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-MacOSX-arm64.sh

Download and install from this link, then run the 3 commands:

chmod +x ~/Downloads/Miniforge3-MacOSX-arm64.sh
sh ~/Downloads/Miniforge3-MacOSX-arm64.sh
source ~/miniforge3/bin/activate

Don’t forget to restart the terminal

You should see (base) appear on the left side of your terminal, like this:

3) Create Environment

Go to a directory and create a test folder. This can be anywhere.

mkdir test
cd test

Now create an environment here:

conda create --prefix ./env python=3.8
conda activate ./env

4) Install Tensorflow and its dependencies

conda install -c apple tensorflow-deps

python -m pip install tensorflow-deps

python -m pip install tensorflow-macos

python -m pip install tensorflow-metal

5) Confirm Tensorflow is working

Now install jupyter, pandas, numpy, matplotlib, scikit-learn in one command!

conda install jupyter pandas numpy matplotlib scikit-learn

Now open jupyter notebook by typing this command in terminal:

jupyter notebook

Run this code in jupyter to confirm successful installation:

import numpy as np
import pandas as pd
import sklearn
import tensorflow as tf
import matplotlib.pyplot as plt
print(tf.config.list_physical_devices())

Confirm your Tensorflow is able to access GPU, it should be printed.