ml-docker

A customized docker container including common machine learning libraries and dependencies.

Acknowledgements

This dockerized ml framework is by no means my sole work. Large shoutout to @waleedka for providing a baseline for combinging the necessary dependencies in an accessible way.

About

This docker build contains common machine learning libraries all built to work with Python 3.5 The libraries included:

• Ubuntu 16.04 LTS
• Python 3.5.2
• Tensorflow 1.6.0
• Keras 2.1.5
• PyTorch 0.3.1
• OpenCV 3.4.1
• Jupyter Notebook
• iPython
• Numpy, Scipy, Scikit Learn, Scikit Image, Pandas, Matplotlib, Pillow
• IBM CPLEX
• Networkx

Requirements

The usage of this dockerfile clearly requires you to have Docker installed: Docker--this can be installed on Mac OSX, Windows or any flavor of Linux. This container download also requires 3.47 GB of disk storage.

Usage

Run in your command line of choice to pull the pre-compiled project:

docker pull mahaloz/ml-docker

Once downloaded, run this to enter the command line of the docker:

docker run -it mahaloz/ml-docker

once inside, you will be in Ubuntu 16.04 bash.

To load a filesystem (volume), use the -v command like so:

docker run -it -v /dir/location/on/local/:/mount/loc/on/docker/ mahaloz/ml-docker 

To use a GUI and load a volume (like what you use in matplot):

docker run --rm -it \
   --user=$(id -u) \
   --env="DISPLAY" \
   --workdir=/ml-docker \
   --volume="$PWD":/ml-docker \
   --volume="/etc/group:/etc/group:ro" \
   --volume="/etc/passwd:/etc/passwd:ro" \
   --volume="/etc/shadow:/etc/shadow:ro" \
   --volume="/etc/sudoers.d:/etc/sudoers.d:ro" \
   --volume="/tmp/.X11-unix:/tmp/.X11-unix:rw" \
   mahaloz/ml-docker

TODO

Need an updated GPU version, consider brining over the No'a project.