🐾 Unitree Go2 - RL Controller (Fast Forward Demo)

This project showcases a fast-forward version of the Reinforcement Learning (RL) controller/policy deployed on a Unitree Go2 quadruped robot.

Watch the demo on YouTube:

📝 Project Overview

• 🎯 Objective: Learn a locomotion controller using RL for the Unitree Go2 and deploy it on real hardware.
• 🚀 Deployment: The trained policy is executed on the Go2 robot in real time.
• ⏩ Fast-forwarded Playback: The video demo is sped up to highlight gait stability, trajectory following, and body dynamics.
• ⚙️ Frameworks: Training was done using Stable-Baselines3, simulation via MuJoCo, and deployment integrated with the Unitree SDK.

📂 Repository Structure

📁 controller/          # Policy implementation and deployment interface
📁 training/            # RL training scripts and environment definitions
📁 scripts/             # Launch, logging, and post-processing tools
📄 README.md            # You're here!