https://github.com/stephane-caron/rl-for-legged-robots
Slides for an introduction to RL in legged robotics
https://github.com/stephane-caron/rl-for-legged-robots
Last synced: 4 months ago
JSON representation
Slides for an introduction to RL in legged robotics
- Host: GitHub
- URL: https://github.com/stephane-caron/rl-for-legged-robots
- Owner: stephane-caron
- License: cc-by-4.0
- Created: 2023-10-16T09:33:59.000Z (over 1 year ago)
- Default Branch: main
- Last Pushed: 2025-01-14T08:52:29.000Z (5 months ago)
- Last Synced: 2025-02-24T10:46:44.674Z (4 months ago)
- Language: TeX
- Homepage:
- Size: 16.4 MB
- Stars: 10
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Reinforcement learning for legged robots
This is a crash course on applying reinforcement learning to train policies that balance real legged robots. We first review the necessary basics: partially-observable Markov decision processes, value functions, the goal of reinforcement learning. We then focus on policy optimization: REINFORCE, policy gradient and proximal policy optimization (PPO). After some practical advice on training with PPO, we finally focus on techniques to train real-robot policies from simulation data: domain randomization, simulation augmentation and reward shaping.
## Usage
- Build slides: ``make``
- Rebuild slides on source updates: ``make watch``
## History
This lecture has been given in the following classes:
- *Robotics* at [MVA](https://www.master-mva.com/cours/robotics/) (Fall 2023, Fall 2024)
- *Introduction to Robotics* (part 2) at Mines de Paris (Fall 2023, Fall 2024)
- *Planification de mouvement en robotique et en animation graphique* at [ENS Paris](https://www.ens.psl.eu/) (Fall 2023, Fall 2024)