Dissertation Project: Continual Reinforcement Learning

Parking environment adapted from HighwayEnv (date accessed: 12/05/2024)

PPO adapted from implementation by Wenlong Wang (date accessed: 25/06/2024)

Dependencies

1. Create and activate virtual environment:
   • With conda:
     • conda create -n thesis python=3.10
     • conda activate thesis
   • Alternatively with venv:
     • python -m venv .venv
     • source .venv/bin/activate
2. Install dependencies: pip install -r requirements.txt

Running Experiments

1. Create configuration JSON file
2. Run python train_eval.py --config PATH_TO_CONFIG