models: PPO model, graph cnn model params.py: setting parameters of environment and neural network main.py: training model file, including RL method(PPO) tool.py: some supplementary functions utils.py: some action functions
Actor-Critic algorithms:https://papers.nips.cc/paper/1999/file/6449f44a102fde848669bdd9eb6b76fa-Paper.pdf PPO method:https://arxiv.org/abs/1707.06347