Objectives
First benchmark tooling. Further appetite will come with eating...

Todos

• Log current commit/git diff
• ability to run with cProfile / and document who to visualise the profils with snakeviz
• cli/notebook for matrix training/evaluation (see Flatland policy evaluation and training cli. #74), write logs of one "run" to folder and upload to S3
• cli/notebook for generating a matrix of results from one run (e.g. comparing average epoch time across different algorithms/policies)
• cli/notebook for comparing different results (e.g. timeline of average epoch times before and after a certain refactoring)

Minimal testing requirements
Think about what behavior shall be covered by a test (qualitatively describe each test).

Additional context
Preparation for #59
Depends on #77 #74