This repository is used for the final project of the Deep Learning course at the Technical University of Denmark.
It contains an implementation of the state-of-the-art reinforcement algorithm SAC as well as a hyper-parameter tuning bench.
Further information on the algorithm, the implementation, and the experiments conducted for this project see the notebook evaluation.ipynb and the report in this repository.
To run the program one needs to run the main.py file with a GPU. It's is possible to change hyperparameters by defining flags.