Dopamine: A framework for flexible value-based reinforcement learning research

Dopamine is a framework for flexible, value-based, reinforcement learning research. It was originally written in TensorFlow, but now all agents have been implemented in JAX.

You can read more about it in our github page and in our white paper.

Original Google AI blogpost.

We have a website where you can easily compare the performance of all the Dopamine agents, which I find really useful:

.

We also provide a set of Colaboratory notebooks that really help understand the framework:

• Create an agent by either subclassing DQN or creating a new agent from scratch
• Train DQN and C51 on the Cartpole environment
• Load and visualize the logs data produced by Dopamine
• Visualize a trained agent using the visualization utilities provided with Dopamine
• Visualize a trained JAX agent using the visualization utilities provided with Dopamine
• Download and visualize different agents with Tensorboard