Feature/implement ray

[Sam-v6]

What does this PR do?

Adding Ray Tune for hyper-parameter optimization

What are some specific details on the changes?

• Paths
  • Replacing os.path.join() with path objects
  • Replacing environment.sh with paths.py which bring in PROJECT_ROOT and SAVE_DIR
• Ray Tune
  • Bringing in Ray Tune with MLflow callback and checkpoint dir
  • Allowing for optional tune run to do hyper-param optimization with --tune
  • Made training more generic to allow HPO to run over top of it if desired
• Miscl
  • Making timing decorator off by default, can be called in training with --profile
  • Getting rid of un-used files
  • Adding generate week 9 predictions to the pipeline

How has this been tested?

Tune is not tested in pipeline, need to implement conditional checks later

Important Notes

None

Checklist

• Tests have been added or updated to cover the changes.
• Documentation has been updated (if necessary)