Use ML flow for experiment tracking

https://mlflow.org/
Probably not necessary for this project, but since this is fairly simple from an ML perspective, it is a nice example for getting to grips with using MLops tools like MLflow. A lot of this is already in place with how this has been set up, this transitions to using a proper tool for structuring experiments rather than my hacked together tool.

Here is a tutorial
https://databricks.com/discover/managing-machine-learning-lifecycle/mlflow-projects-and-models