Purpose

• Run tensorflow and Jupyter notebook server, probably continuum.io anaconda.
• Exists as docker-compose even though this is single node to make multi-node easier in future.
• Guide: tensorflow (python 3.x) https://hub.docker.com/r/tensorflow/tensorflow/

Features

• Mounts named volume into /tf/notebooks to store files across restarts
• Mounts file share ~/Documents/GitHub into /tf/notebooks/GitHub
• The Jupyter web server is located at http://localhost:8888/
  • The url with the secret token will show up in the logs of the window you started this with
• Open shell command prompt with docker exec -it tensorflow-py3-jupyter /bin/bash
  • This may not work from a GitHub Bash prompt on Windows

Docker Management

See README.md in the root of this repo for interesting docker commands.

IDE Integration

No IDE integrations specific to this project are called out at this time.