MatCom Machine Learning image

Base machine learning images and environment with CPU and GPU support.

This repository contains two images:

• matcomuh/ml:cpu is a basic machine learning image with several popular ML tools.
• matcomuh/hub:cpu is a fully-functional JupyterHub on top of the basic ML image.

Also, the same images with GPU support:

• matcomuh/ml:gpu
• matcomuh/hub:gpu

Basic ML usage

If you just want to hack machine learning on your own, you can use the basic image. Clone this repository and run:

docker-compose up ml

In localhost:8888 you will find an instance of JupyterLab. The notebooks are stored in the local notebooks folder.

JupyterHub

If you need a more advanced multi-user JupyterHub scenario, then run:

docker-compose up hub

In localhost:8000 you will find an instance of JupyterHub.

• The default user is admin with password admin.
• The file hub/config.py contains the configuration file for this instance.

Users are by default added to the system, and their data folders are mounted in a docker volume. Hence when the container is re-created the data and users will still be there.

NOTE: New users are created by default with the same username as password. When the container is destroyed and re-created, password changes are not saved for now.

Running the GPU version

By default the CPU version of the services are run. If you want to try the GPU version, you will need nvidia-docker2 installed, and suitable NVIDIA drivers for your box.

With all prerequisites, you are ready to run the GPU version of the services:

docker-compose -f docker-compose.yml -f docker-compose.gpu.yml up [ml|hub]

If you are gonna be running GPU all the time, consider creating a docker-compose.override.yml link to simplify things:

ln -s docker-compose.gpu.yml docker-compose.override.yml

Then just running docker-compose up as usual will automatically use the GPU version of the services.

What's included

• Jupyter Notebook / Lab / Hub
• Tensorflow (1.12.0)
• Keras (2.1.6-tf) (see note)
• Scikit-learn (0.20)
  • hmmlearn
  • sklearn-crfsuite
  • seqlearn
• Flask & Flask-RESTful
• Gensim
• Graphviz
• NLTK
• owlready (1 & 2)
• Spacy
  • (en and es corpora)
  • neuralcoref

Plus small utilities such as psutils. Take a look at the requirements.txt file.

NOTE: To use keras, you have to import it as from tensorflow import keras.

Contributors:

• Alejandro Piad
• Hian Cañizares

License & Contributions

All contributions are appreciated! Licensed under MIT. Make sure to add your name to the previous list.

MIT License

Copyright (c) 2018 Faculty of Math and Computer Science, University of Havana

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.