Model & Data Management

[lapplislazuli]

At the moment (v1.0), the model is hard-baked into the Container.
But the repository does not contain a model - so we have to provide one somewhere visible for download.

I think we should have the model mounted in using docker volumes and put one of the models into the reproduction package.

Same goes with a small and a "real" dataset we use.

• Provide Model for easy download
• Provide Data for easy download
• Change Docker Container to take Model per Volume
• Adjust Readme / docs
• Change Experiment Creator to cover volumes