2016 agent code and open-source learning environment framework

[Daniel Ricks]

We have created a learning environment framework for text-based adventure games (a la ALE for Atari games), and we've open-sourced it. It has 50 games for testing, and we've tried to make it as accessible as possible. If you have questions, feel free to ask.

We have also open-sourced our winning agent from last year, but we haven't added comments or explanations. Making that code more accessible isn't at the top of our priority list, but it is currently available for perusal.

Our paper, which was recently accepted by IJCAI, is available here: https://arxiv.org/abs/1703.03429

A link to our learning environment framework: https://github.com/danielricks/autoplay

Our agent: https://github.com/danielricks/BYU-Agent-2016