1. Learn
    2. /
  2. Courses
    3. /
  3. Deep Reinforcement Learning in Python
Connected

Exercise

Instantiating the Q-Network

Now that you defined its architecture, you are ready to instantiate the actual network that your agent will be using, as well as its optimizer. The Lunar Lander environment has a state space of dimension 8, and an action space of dimension 4 (corresponding to 0: do nothing, 1: left thruster, 2: main engine, 3: right thruster).

The QNetwork class from the previous exercise is available to you.

Instructions

100 XP
  • Instantiate a Q Network for the Lunar Lander environment.
  • Define the Adam optimizer for the neural network, specifying a learning rate of 0.0001.