A self-contained backend-independent version of FLOW for (deep) reinforcement learning in a network environment.
This is a minimized version of FLOW (https://github.com/berkeleyflow/flow) stack. It allows users to define a custom environment backend and a custom agent backend. The current version supports SUMO (https://github.com/eclipse/sumo) as the environment backend and tensorflow (https://github.com/tensorflow) as the agent backend.
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.