Setup Overview

• The setup here is a little complicated due to the limitations of not having our own GPU server (for now).
• Since we don't have our own GPUs (yet), we are using Snellius to run the automl pipelines. Now, Snellius does not really want us to do that, hence not allowing for cron jobs. To overcome this, we use the OpenML test server.

Components

• Therefore to deploy the solution, we need to configure multiple things in multiple places. This will be explained in this section.
• Look at snellius first, then test server