Menghan Wang

Menghan Wang

PhD candidate @ HKU

Set Up Environment for Deep Learning on Mac M1

2 minute read

Published:

This article introduce steps to set up the Python environment for deep learning on Mac M1. We will cover

Before we start, if you have already installed Anaconda or other conda environment, you may want to uninstall them first using anaconda-clean.

conda install anaconda-clean
anaconda-clean --yes

To remove the entire directory, try following.

rm -rf ~/anaconda3
rm -rf ~/.anaconda_backup 
rm -rf ~/opt/anaconda3
rm -rf ~/miniforge3

Finally, we need to remove anaconda path in .bash_profile. open .bash_profile from terminal, delete the line export PATH="/Users/your_username/anaconda3/bin:$PATH". If you don’t have permission to write to it, use sudo chown your_username ~/.bash_profile to change the owner of the file, and try again.

Now we could start the main course.

Install Miniforge 3

Download miniforge3 from here. Choose Miniforge3-MacOSX-arm64，and save to Downloads.

miniforge3

Run the command below to install. Follow the default option during installation.

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

You could also use homebrew brew install miniforge.

After you have successfully install Miniforge 3, restart the terminal and move to the next step.

Install Tensorflow

Before installing Tensorflow, we need to create a conda environment for it.

conda create -n tensorflow python==3.9
conda activate tensorflow

Note that tensorflow after conda create -n is the name for this new environment, feel free to use any name you like. I realized at the end that I put Tensorflow and Transformers under one environment and named it “tensorflow”. You could always rename it using conda rename -n old_name -d new_name.

Now we can install tensorflow.

conda install -c apple tensorflow-deps  
python -m pip install tensorflow-macos
python -m pip install tensorflow-metal

To check if the installation is successful, run conda env list, you will find two environments, base and tensorflow. You could also check available kernel by jupyter kernelspec list.

conda_env

Next, let’s install some commonly-used packages. Note that the first two lines set conda-forge as the highest priority channel.

conda config --add channels conda-forge
conda config --set channel_priority strict
conda install jupyter jupyterlab pandas matplotlib seaborn nltk spacy beautifulsoup4 scikit-learn

Finally, open the jupyter lab to check your Tensorflow version and device. You can now try Tensorflow on your mac!

source ~/miniforge3/bin/activate
conda activate tensorflow
jupyter lab

jupyter_notebook

Install Transformers

The final piece we need is Transformers. Before that, we should install PyTorch and Rust.

pip install -U --pre torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/nightly/cpu
curl — proto ‘=https’ — tlsv1.2 -sSf https://sh.rustup.rs | sh

Now you can run pip install transformers to get the job done. Enjoy your journey!


Reference

Tags:

You May Also Enjoy

Synthetic DID and its implementation in Stata

9 minute read

Published:

As one can tell from the name, synthetic DID is a combination of synthetic control and DID. One main challenge of DID is the validation of parallel trend assumption, as it not always holds in the data that treated and control groups have identical trend or characteristics prior to the treatment event. Built on the method of synthetic control, synthetic DID loosens the needs for parallel assumption by constructing matched control units.

OCR Tools in Python: Azure Vision and Tesseract

7 minute read

Published:

Optical Character Recognition (OCR) is a technology that enables computers to read text from images. OCR tools have become increasingly popular due to their ability to automate data entry and extract information from scanned documents. In this blog post, I will cover the usage of two OCR tools in Python:

Accelerating Data Processing in Python with Parallelism

4 minute read

Published:

Parallel data processing is a technique used to speed up data processing tasks by dividing the data into smaller chunks and processing them simultaneously on multiple processors. It could be useful as long as do not need to be implemented sequentially. In contrast to threading, which is suitable for tasks that are I/O bounded but not tasks that are CPU bounded, multiprocessing runs on multiple cores and helps when tasks are CPU bounded. Python provides several ways to implement parallel data processing.