elipugh/euchre

Deep Q Learning for Euchre

RLCard: A Toolkit for Reinforcement Learning in Card Games

This is a fork of rlcard with added Euchre environments and updated PyTorch agents so you can use RLCard to train agents for Euchre! Most of the Euchre environment comes from this repo, but it has been heavily modified to work with DQN and NFSP agents.

This is also a project for the Stanford class AA228, Decision Making Under Uncertainty. You can check out the final report here

Examples

All examples are located in Colab notebooks. Check them:

• Deep Q learning
• Deep Q learning with pretraining
• NFSP
• NFSP with pretraining