Configuring a ML Server

This is a run-down of how we set up a server for machine learning, along with information on how to connect and use the server.

One of the difficulties is that the network our server is on disallows inbound ssh connections - meaning that we can ssh from the server to external servers, but not externally into our server. We can get around this by using ssh tunneling.

How to Connect

To connect, you will need to set up an ssh tunnel.

After that, you should navigate to localhost:8156 and see a jupyter notebook.

Setting up the server

Install your favorite linux distribution. For this project, we ended up using ubuntu as it was the simplest to set up.

Install docker. We used a deep-learning docker image you can find here.

Next we wrote a custom makefile to bring up the docker image and expose the jupyter-notebook ports.