Warning
These instructions are still mostly untested. Please open an issue if you run into any problems. See the release of FastRLAP for general setup instructions.
On the robot side, install jaxrl5 and put the offroad-learning directory into ~/ros_ws/src. On the training side, install everything normally (ROS is not explicitly required on the training computer).
Once you've got everything set up, you'll need to collect a set of goal checkpoints defining your course:
rosrun offroad_learning goal_graph_recorder_node _fixed_frame:="map"Drive the course, pressing the B button at each goal location. This will print out a set of goals, which you can put in offroad_learning/config/goals/<environment name>/goals.yaml.
To launch robot-side inference, you can use roslaunch offroad_bringup real_inference.launch. On the training computer, run python offroad_learning/src/offroad_learning/training/training.py.