Embedded Files
Shifu is a lightweight IsaacGym wrapper for robot learning. It trains robots in a massively parallel way like IsaacGym, but also it is simple to build, highly readable, and easily customized. This project aim to help robot learning researchers enjoy the incredible IsaacGym’s few minutes policy training without complex engineering
Lightweight
Lightweight
ShifuVecEnv is an object-orientated IsaacGym wrapper. The Pytorch tensors in the IsaacGym are directly managed by the Units(robot, objects, and sensors ) objects.
Page updated
Google Sites
Report abuse