Hands-On Reinforcement Learning for Games

Exercises

As we progress through the book, I hope you can see the value of performing these additional hands-on exercises. Learning how to tune hyperparameters will be essential in building DRL models that can tackle difficult environments. Use the following exercises to reinforce your learning of the material:

1. Modify the batch_size, inputs, hidden, and outputs hyperparameters from Chapter_6_1.py and see what effect these have on the output loss.
2. Alter the number of training iterations in the Chapter_6_1.py example in conjunction with other hyperparameters in order to evaluate the impact this has on training.
3. Modify the batch_size, inputs, hidden , and outputs hyperparameters from Chapter_6_2.py, and see what effect these have on the output ...