ml-agents/docs

Unity ML-Agents Toolkit Documentation

Installation & Set-up

• Installation
  • Background: Jupyter Notebooks
  • Using Virtual Environment (Recommended)
  • Docker Set-up
• Basic Guide

Getting Started

• ML-Agents Toolkit Overview
  • Background: Unity
  • Background: Machine Learning
  • Background: TensorFlow
• Getting Started with the 3D Balance Ball Environment
• Example Environments

Creating Learning Environments

• Making a New Learning Environment
• Designing a Learning Environment
  • Agents
  • Academy
  • Brains: Player, Heuristic, Learning
• Learning Environment Best Practices

Advanced Usage

• Using the Monitor
• Using the Video Recorder
• Using an Executable Environment
• Creating Custom Protobuf Messages

Training

• Training ML-Agents
• Using TensorBoard to Observe Training
• Training with Proximal Policy Optimization
• Training with Soft Actor-Critic

Advanced Training Methods

• Training with Curriculum Learning
• Training with Imitation Learning
• Training with LSTM
• Training Generalized Reinforcement Learning Agents
• Training Using Concurrent Unity Instances

Cloud Training (Unsupported)

Here are the cloud training set-up guides for Azure and AWS. We no longer use them ourselves and so they may not be work correctly. We've decided to keep them up just in case they are helpful to you. If you do try them out, please share your experience with us, and feel free to submit a PR with any necessary changes.

• Training on the Cloud with Amazon Web Services
• Training on the Cloud with Microsoft Azure

Inference

• Unity Inference Engine

Help

• Migrating from earlier versions of ML-Agents
• Frequently Asked Questions
• ML-Agents Glossary
• Limitations

API Docs

• API Reference
• How to use the Python API
• Wrapping Learning Environment as a Gym (+Baselines/Dopamine Integration)
• Creating custom protobuf messages