- Install environment:
git clone https://github.com/ireneb97/RA_project.git
cd RA_project
- Install Lydia (before you need to install Docker):
docker pull whitemech/lydia:latest
echo '#!/usr/bin/env sh' > lydia
echo 'docker run -v$(pwd):/home/default whitemech/lydia lydia "$@"' >> lydia
sudo chmod u+x lydia
sudo mv lydia /usr/local/bin/
- Create and initialize environment:
python3 -m venv ./venv
source venv/bin/activate
- Install dependencies:
pip install -r requirements.txt
In folder config
are stored some configurations we have used. We suggest to not to change those files as they already store the best values for each map and algorithm pair.
You can run one of them (e.g. config1.cfg) by running the command:
python3 main.py --config_file config1.cfg
In folder model
we saved our trained agents, you can run one of them by using this command:
python3 main.py --trained_model_path model/ppo
- You can read in details about this project here, inside our report.
- You can see our video presentation here.
- You can see the slides we used in our presentation here.
- Irene Bondanza (bondanza.1747677@studenti.uniroma1.it)
- Matteo Emanuele (emanuele.1912588@studenti.uniroma1.it)
- Pietro Manganelli Conforti (manganelliconforti.1754825@studenti.uniroma1.it)