Right now this is just a scratchpad.  I'm converting my messy ML lab environment to something with better MLOps standards

Broadly, the goals are:
- Local dev environment that's as similar as possible to production
- A managed Jupyter Lab environment for experimentation
- Do something sensible with features and feature extraction
- A good mockup of an operationalized model
- CI/CD pipeline with local access to test/verification


## Requirements

- docker
- https://github.com/NVIDIA/nvidia-container-toolkit