The codebase replicating the results of the "Ensembling Diverse Policies Improves Generalizability of Reinforcement Learning Algorithms in Continuous Control Tasks" paper and similarly titled thesis work.
Python 3.9 is recommended
pip install -r requirements.txt
pip install citylearn==1.3.6 --no-deps
python train.py --wandb-entity <wandb team/user name>