This is the code for paper "DyPS: Dynamic Parameter Sharing in Multi-Agent Reinforcement Learning for Spatio-Temporal Resource Allocation "
conda create -n DyPs python=3.9
conda activate CoTa
pip install -r requirements.txt- Command to run our method
cd run
python run_lstm_hrl.py- Command for baselines
cd run
python run_hrl.py # Our method without lstm
python run_cvae.py # Our method withour hierarchical structure
python PS_noid.py # Naive parameter sharing - You can visualize the learning curves by tensorboard.
tensorboard --logdir logs- You can visualize the city map and demand-supply heat map by following jupyters.
plot/grid_map.ipynb- training curve of ride-hailing (2) scenario