Student project which trains the CartPole-v0 environment of OpenAI, using a actor-critic reinforcement learning method.
The notebook was created in Google Colab therefore there might be problem running it on a local machine. A link to my Colab Notebook is in the .txt file. The gif shows the finished product of the training.
I found this Tutorial on Tensorflow https://www.tensorflow.org/tutorials/reinforcement_learning/actor_critic and wanted to rebuild it myself as I was already fascinated by a Generative Adversarial Network example shown in our lecture. I especially liked the Generator-Discriminator connection, which can also be found to some degree here with the Actor-Critic connection.