TensorFlow macOS Setup Guide: Complete Installation Tutorial

Install Homebrew Package Manager

Homebrew is the essential package manager for macOS, functioning similarly to apt-get on Ubuntu or yum on Red Hat Linux. This tool simplifies software installation and dependency management on your Mac. Homebrew requires Xcode to be installed, but the installation command below will automatically install Xcode if it’s not already present on your system.

/usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"

Install Python 3

While macOS comes with Python 2 installed by default, the development community has migrated to Python 3 for modern applications and machine learning frameworks. TensorFlow requires Python 3 for optimal performance and access to the latest features. Installing Python 3 through Homebrew ensures you have the most recent stable version with proper dependency management.

brew install python3

Install pip3 Package Manager

pip is Python’s package installer, essential for managing Python libraries and dependencies. There are two versions available: pip (for Python 2) and pip3 (for Python 3). Since we’re working with Python 3 for TensorFlow installation, we’ll use pip3 to ensure compatibility and access to the latest package versions.

Understanding pip vs pip3

• pip – Installs packages for Python 2 (legacy version)
• pip3 – Installs packages for Python 3 (current standard)

Download and install pip3 using the following command:

curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py

Install TensorFlow

With all prerequisites installed, you can now install TensorFlow. The installation command varies depending on your Python version. Since we’ve set up Python 3, we’ll use pip3 for the installation to ensure compatibility with the latest TensorFlow features and optimizations.

For Python 3 (Recommended)

pip3 install tensorflow

For Python 2 (Legacy)

If you’re still using Python 2 (not recommended for new projects), use this command:

pip install tensorflow

Quick Installation Commands

For convenience, here are all the installation commands consolidated into a single code block. You can copy and paste these commands to run the complete installation process:

/usr/bin/ruby -e "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/master/install)"
brew install python3
curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
pip3 install tensorflow

Testing Your TensorFlow Installation

To verify that your TensorFlow installation is working correctly, you can test it with a practical machine learning example. We recommend trying out this stock predictor written in Python, which demonstrates TensorFlow’s neural network capabilities:

GitHub Repository: https://github.com/sebastianheinz/stockprediction

This repository contains both the neural network code and training data. To run the stock predictor, you’ll need to install additional Python packages:

pip3 install matplotlib
pip3 install pandas
pip3 install numpy
pip3 install sklearn
pip3 install scipy

For a detailed explanation of the stock market predictor implementation, you can read this comprehensive guide: A Simple Deep Learning Model for Stock Price Prediction Using TensorFlow

Troubleshooting Common Issues

macOS High Sierra Version Compatibility

If you encounter a runtime warning similar to “RuntimeWarning: compiletime version 3.5 of module ‘tensorflow.python.framework.fast_tensor_util’ does not match runtime version 3.6” while running macOS High Sierra, this indicates a version mismatch between the compiled TensorFlow binary and your Python runtime.

To resolve this issue, install the macOS High Sierra compatible version:

pip3 install --ignore-installed --upgrade "https://github.com/lakshayg/tensorflow-build/raw/master/tensorflow-1.4.0-cp36-cp36m-macosx_10_12_x86_64.whl"