Robot Learning: A Tutorial

This repository contains the source code for the "Robot Learning: A Tutorial" report. This tutorial covers many of the most pressing aspects in modern robot learning, and provides practice examples using lerobot, the robot-learning library developed by Hugging Face.

You’re more than welcome to contribute to the next edition of the tutorial! Simply open an issue, tag @fracapuano, and start a discussion about the scope and content you’d like to add. Check out CONTRIBUTING.md for more details 😊 All merged pull requests will receive public acknowledgment in the main body of the tutorial. Items marked with an empty [ ] in the following Table of Contents are open for community contribution!

Table of Contents

1. Introduction

• [x] 1.1 lerobot Dataset
  • [x] 1.1.1 The dataset class design
• [x] 1.2 Code Example: Batching a (Streaming) Dataset
• [x] 1.3 Code Example: Collecting Data

2. Classical Robotics

• [x] 2.1 Explicit and Implicit Models
• [x] 2.2 Different Types of Motion
• [x] 2.3 Example: Planar Manipulation
  • [x] 2.3.1 Adding Feedback Loops
• [x] 2.4 Limitations of Dynamics-based Robotics

3. Robot (Reinforcement) Learning

• [x] 3.1 A (Concise) Introduction to RL
• [x] 3.2 Real-world RL for Robotics
• [x] 3.3 Code Example: Real-world RL
• [x] 3.4 Limitations of RL in Real-World Robotics: Simulators and Reward Design

4. Robot (Imitation) Learning

• [x] 4.1 A (Concise) Introduction to Generative Models
  • [x] 4.1.1 Variational Auto-Encoders
  • [x] 4.1.2 Diffusion Models
  • [x] 4.1.3 Flow Matching
• [x] 4.2 Action Chunking with Transformers
  • [x] 4.2.1 Code Example: Training and Using ACT in Practice
• [x] 4.3 Diffusion Policy
  • [x] 4.3.1 Code Example: Training and Using Diffusion Policies in Practice
• [x] 4.4 Optimized Inference
  • [x] 4.4.1 Code Example: Using Async Inference

5. Generalist Robot Policies

• [x] 5.1 Preliminaries: Models and Data
• [x] 5.2 Modern VLAs
  • [x] 5.2.1 VLMs for VLAs
• [x] 5.3 PI0
  • [x] 5.3.1 Code Example: Using PI0
• [x] 5.4 SmolVLA
  • [x] 5.4.1 Code Example: Using SmolVLA
  • [ ] 5.4.2 Code Example: Training SmolVLA
• [ ] 5.5 GR00T (1/2)
  • [ ] 5.5.1 Code Example: Using GR00T
• [ ] 5.6 PI05
  • [ ] 5.6.1 Code Example: Using PI05
• [ ] Large-scale datasets
  • [ ] Open-X
  • [ ] DROID
  • [ ] BEHAVIOR

6. Some Emerging Directions in Robot Learning

• [ ] 6.1 Post-training VLAs
  • [ ] 6.1.1 From Imitation to Refinement
  • [ ] 6.1.2 EXPO
• [ ] 6.2 World Models for robotics
  • [ ] 6.2.1 Cosmos
  • [ ] 6.2.2 World Models (1X)
  • [ ] 6.2.3 Sima and Genie 1

7. Conclusions

• [x] 7.1 Conclusions

License

The written content of this book is licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

All source code examples in the snippets/ directory are licensed under the MIT License.