This repository adapts the walk-these-ways project to train and deploy reinforcement learning policies on Unitree Go2 robots. It uses Isaac Gym and PyTorch for training locomotion policies through reinforcement learning, and interfaces with unitree-sdk2 for deployment on the actual robot hardware. The project enables sim-to-real transfer of learned gait-conditioned agility behaviors.