The codebase accompyning the paper "Optimistic Online Non-stochastic Control via OFTRL".
-
To run an experiment, activate a scenario in env.py, and then execute main.py.
-
The results will be stored in "./results", running regret_plotting.py will then produce the plots.