On this page
Get to grips on the latest AI power in Unity
The Unity Machine Learning Agents SDK (ML-Agents), currently in beta, is an open-source Unity plugin that enables games and simulations to serve as environments for training intelligent agents. Agents can be trained using reinforcement learning, imitation learning, neuroevolution, or other machine learning methods through a simple-to-use Python API. We collected together the latest resources for you to get started with the SDK.
From GDC 2018: Unity ML-Agents: what it is, how it works, how to use it
This is a helpful and very accessible session to start learning about ML-Agents. Danny Lange, VP of Machine Learning and AI at Unity gives the overview on the most exciting advances that AI will bring over the next 12 months. Then, Arthur Juliani, Lead Engineer on ML-Agents explains the workflow in Unity, how the plugin works with a Unity project, the current training methods available, and the latest and upcoming features.
An in-depth intro to ML-Agents
v0.1: An introduction to ML-Agents
Similar to his GDC session, this blog post by Arthur is a thorough and easy-to-understand introduction to the tech behind ML-Agents, and how you can use the SDK.
v0.2: Curriculum Learning, new environments, and more
In this post, Arthur explains the new features added in v0.2
v0.3: Imitation Learning, feedback-driven features, and more
Another helpful post by Arthur explaining new additions such as Imitation Learning, Multi-Brain Training, On-Demand Decision-Making, and Memory-Enhanced Agents.
ML-Agents in a real game: a beginner’s guide
Unity evangelist Alessia Nigretti explains what her and her fellow evangelist Ciro Continisio learned while making their demo that uses ML-Agents.
Exploring the concepts of Machine Learning and AI
In these two blog posts, Arthur takes a more exploratory approach, explaining some of the basic concepts of Machine Learning. As he writes "these (posts) are a creative space to explore all these emerging developments"