Check list for various use cases of the repo

This is a check list of what would need to be done for various use cases of the repo. This list doesn't include what we would need for using Emote in production for our projects.

Basics:
- [ ] #18
- [ ] #27
- [x] #19
- [x] #20

Open-source:
- [ ] #21
- [ ] #22

Research for Arkadia:
- [ ] #23
- [ ] #24
- [ ] #25
- [ ] #26

Nice to haves:
- [ ] More advanced training/architecture stuff (Spectral normalization, Observation normalization, Reward normalization, Q-Value/Advantage normalization, ...)
- [ ] More network types maybe?
- [ ] Support for Multi-Discrete action spaces
- [ ] CURL or other vision methods
- [ ] Hyper parameter tuning