/Deep-Q-Learning-Implementation-using-Tensorflow-2.0

This is my DQN implementation with Experience Replay.

Primary LanguageJupyter Notebook

Deep-Q-Learning-Implementation-using-Tensorflow-2.0

This is my DQN implementation with Experience Replay.

In this implementation, I have used Pacman environment of Atari OpenAI Gym. I've trained the DQN for an hour. It yielded reasonable learning results as the number of episodes increased.

If the network is trained for longer hours it will yield great results.

We can implement various OpenAI gym environments in the DQN.

References:

  • I have referred to simoninthomas DQN repo.
  • I also referred to Tensorflow 2.0 documentation.