ROSject of the week: train a DeepQNN to control a cartpole with Gazebo sim, ROS control and OpenAI

In this rosject we show you how to build the whole training system for a cartpole simulated in Gazebo and controlled with ROS using OpenAI framework.

You can get the full ROS training code, documentation in the form of Jupyter notebook, openai_ros package and cartpole Gazebo simulation here:

1 Like