Model modules

[Veldhoen]

• [ ] Speaker diarisation
• [ ] Speech Recognition

Module(s) where you actually run your model, possible preprocessing, etc etc

• Add unit tests for relevant bits