MineRL: Towards AI in Minecraft

Welcome to documentation for the MineRL project and its related repositories and components!

What is MineRL

MineRL is a research project started at Carnegie Mellon University aimed at developing various aspects of artificial intelligence within Minecraft. In short MineRL consists of several major components:

• MineRL-v0 Dataset – One of the largest imitation learning datasets with over 60 million frames of recorded human player data. The dataset includes a set of environments which highlight many of the hardest problems in modern-day Reinforcement Learning: sparse rewards and hierarchical policies.

• minerl – A rich python3 package for doing artificial intelligence research in Minecraft. This includes two major submodules. We develop minerl in our spare time, please consider supporting us on Patreon

  • minerl.env – A growing set of OpenAI Gym environments in Minecraft. These environments leverage a synchronous, stable, and fast fork of Microsoft Malmo called MineRLEnv.

  • minerl.data – The main python module for ext with the MineRL-v0 dataset

Tutorials and Guides

• Installation
• Hello World: Your First Agent
  • Creating an environment
  • Taking actions
  • No-op actions and a better policy
• Sampling The Dataset
  • Moderate Human Demonstrations
• K-means exploration
• Visualizing The Data minerl.viewer
• Interactive Mode minerl.interactor

MineRL Environments

• General Information
• Environment Handlers
  • Environment Handlers
  • Observations
  • Actions
• Basic Environments
• Competition Environments

Notes

• Windows FAQ
  • The freeze_support error (multiprocessing)

MineRL package API Reference

• minerl.env
  • MineRLEnv
  • InstanceManager
• minerl.data
  • MineRLv0

Indices and tables

• Index

• Module Index

• Search Page