https://github.com/mujocolab/mjlab_playground
A collection of tasks built on mjlab
https://github.com/mujocolab/mjlab_playground
humanoid mjlab mujoco quadruped reinforcement-learning robotics sim2real
Last synced: about 1 month ago
JSON representation
A collection of tasks built on mjlab
- Host: GitHub
- URL: https://github.com/mujocolab/mjlab_playground
- Owner: mujocolab
- License: apache-2.0
- Created: 2026-04-03T01:23:22.000Z (2 months ago)
- Default Branch: main
- Last Pushed: 2026-04-15T06:34:29.000Z (about 2 months ago)
- Last Synced: 2026-04-15T07:30:47.817Z (about 2 months ago)
- Topics: humanoid, mjlab, mujoco, quadruped, reinforcement-learning, robotics, sim2real
- Language: Python
- Homepage: https://mujocolab.github.io/mjlab
- Size: 7.56 MB
- Stars: 69
- Watchers: 0
- Forks: 2
- Open Issues: 1
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# mjlab playground
A collection of tasks built with [mjlab](https://github.com/mujocolab/mjlab), starting with ports from [MuJoCo Playground](https://playground.mujoco.org/).
## Tasks
| Task ID | Robot | Description | Preview |
|---------|-------|-------------|---------|
| **Getup** | | | |
| `Mjlab-Getup-Flat-Unitree-Go1` | Unitree Go1 | Fall recovery on flat terrain | 
|
| `Mjlab-Getup-Flat-Booster-T1` | Booster T1 | Fall recovery on flat terrain | 
|
## Getting Started
```bash
git clone https://github.com/mujocolab/mjlab_playground.git && cd mjlab_playground
uv sync
```
Train a task:
```bash
uv run train --num_envs 4096
```
Play back a trained policy:
```bash
uv run play
```
### Getup training
On a single NVIDIA 5090, the Go1 getup task converges in ~2 minutes and T1 in ~8 minutes, but we continue training with a curriculum that progressively tightens action rate, joint velocity, and power penalties to produce smoother, safer policies.
## Citation
If you use this repository in your research, consider citing mjlab:
```bibtex
@misc{zakka2026mjlablightweightframeworkgpuaccelerated,
title={mjlab: A Lightweight Framework for GPU-Accelerated Robot Learning},
author={Kevin Zakka and Qiayuan Liao and Brent Yi and Louis Le Lay and Koushil Sreenath and Pieter Abbeel},
year={2026},
eprint={2601.22074},
archivePrefix={arXiv},
primaryClass={cs.RO},
url={https://arxiv.org/abs/2601.22074},
}
```
## License
This repository is released under an [Apache-2.0 License](LICENSE).