Munchause-RL

PyTorch implementation of the M-DQN algorithm based on the paper Munchause Reinforcement Learning.

For a short introduction check out the Medium Article!

Implementations

Discrete Action Space:

• M-DQN
• M-IQN
• M-FQF

Continuous Action Space:

• M-SAC

Changes to the Paper

Compared to the original algorithm I did some changes:

1. Instead of doing a hard update every 8000 frames I implemented a soft-update. By personal experience this worked better.

Results

Comparison runs between M-DQN and DQN for the CartPole-v0 environment and LunarLander-v2.

Comparison of IQN and M-IQN for LunarLander-v2

Comparison IQN and M-IQN for Breakout