IMRT-DRL I build the gym environment of PortPy to use DDQN to train the agent that can choose the proper beam angles. The repository includes portpy-gym environment building and network traning codes.