Training to land softly

  • Follow the steps that you learned from doing the tutorial to understand how to get your lunar lander to land softly. You will have to tune hyperparameters, train your RL model, and analyze metrics to understand how your agent is performing.

  • Remember that since you are working on the first objective, edit the scenario variable to be scenario = '1' as shown below:

You can pick and choose which scenarios you want to work on. You do not have to work on all of them. You can focus on optimizing a few and choose to work on them in any order. You can even divide and conquer with your teammate if you choose to do so.