This repository contains a Docker template and a scrip to create AI Docker containers with Python, based on Alpine Linux. The containers come with sipy, numpy, matplotlib, scikit-learn and jupyterlab.
To create a new Dockerfile with updated Python packages, simply edit env.sh and execute create_dockerfiles.sh. This will tag and push the Docker files to GitHub, what will trigger automatic builds on Docker Hub. The images are available here and are built with the following settings:
| Branch/Tag | Source | Location | Name |
|---|---|---|---|
| Tag | /^py-\d.\d_ap-\d.\d___np-\d+.\d+.\d+_sp-\d+.\d+.\d+_sk-\d+.\d+.\d+$/ |
/_versioned |
latest |
| Tag | /^py-\d.\d_ap-\d.\d___np-(\d+.\d+.\d+)_sp-(\d+.\d+.\d+)_sk-(\d+.\d+.\d+)$/ |
/_versioned |
np-{\1}_sp-{\2}_sk-{\3} |
| Tag | /^py-(\d.\d)_ap-\d.\d___np-(\d+.\d+.\d+)_sp-(\d+.\d+.\d+)_sk-\d+.\d+.\d+$/ |
/_versioned |
py-{\1}_np-{\2}_sp-{\3} |
| Branch | master |
/_versioned |
versioned |
| Tag | /^py-(\d.\d)_ap-\d.\d___np-(\d+.\d+.\d+)_sp-(\d+.\d+.\d+)_sk-\d+.\d+.\d+$/ |
/_versioned-dev |
py-{\1}_np-{\2}_sp-{\3}-dev |
| Branch | master |
/_versioned-dev |
versioned-dev |
| Branch | master |
/_experimental |
experimental |
| Branch | master |
/_experimental-dev |
experimental-dev |
The image is only 95MB!
$ docker run -it -p 8888:8888 anthropocentricai/ai-pythonOnce you have run this command, point your web browser to http://127.0.0.1:8888 and paste the Jupyter Lab token displayed in the terminal.