Final Project for the for the Fundamentals of ML course winter term 21/22
- Jakob, Erben - jakobus.erben@gmail.com - 3727893
- Ebbo, Kramer - Email einfügen - Mat.-Nr. einfügen
- Vanessa, Zuber - zuber@cl.uni-heidelberg.de - 3386382
- report: insert folder-adress
- agent-code: insert folder-adress
- example replays: insert folder-adress
- run a game: python main.py play
- run game with own agent: python main.py play --my-agent my_agent
- run a game with multiple custom agents: python main.py play --agents my_agent random_agent rule_based_agent peaceful_agent
- for training consider: Use --train N (N specifies, that the first N agents passed by --agents should be trained)
- training example: python main.py play --agents my_agent random_agent rule_based_agent peaceful_agent --train 1 OR python main.py play --my-agent my_agent --train 1
- CARE: Training will automatically stop when the last agent to be trained dies. In order to prevent this use the option: --continue-without-training
- Task (Coin Heaven)
-> correct Command: python main.py play --no-gui --agents own_coin_collector --train 1 --scenario coin-heaven