Metadata-Version: 2.4
Name: rsl-rl-lib
Version: 5.0.0
Summary: Fast and simple RL algorithms implemented in PyTorch
Author-email: Clemens Schwarke <cschwarke@ethz.ch>, Mayank Mittal <mittalma@ethz.ch>, Nikita Rudin <rudinn@ethz.ch>, David Hoeller <holler.david78@gmail.com>
Maintainer-email: Clemens Schwarke <cschwarke@ethz.ch>, Mayank Mittal <mittalma@ethz.ch>
License-Expression: BSD-3-Clause
Project-URL: Homepage, https://github.com/leggedrobotics/rsl_rl
Project-URL: Issues, https://github.com/leggedrobotics/rsl_rl/issues
Keywords: reinforcement-learning,robotics
Classifier: Programming Language :: Python :: 3
Classifier: Operating System :: OS Independent
Requires-Python: >=3.9
Description-Content-Type: text/markdown
License-File: LICENSE
License-File: licenses/dependencies/codespell-license.txt
License-File: licenses/dependencies/furo-license.txt
License-File: licenses/dependencies/gitpython-license.txt
License-File: licenses/dependencies/numpy-license.txt
License-File: licenses/dependencies/onnx-license.txt
License-File: licenses/dependencies/onnxscript-license.txt
License-File: licenses/dependencies/pre-commit-hooks-license.txt
License-File: licenses/dependencies/pre-commit-license.txt
License-File: licenses/dependencies/pyright-license.txt
License-File: licenses/dependencies/ruff-license.txt
License-File: licenses/dependencies/setuptools-license.txt
License-File: licenses/dependencies/sphinx-license.txt
License-File: licenses/dependencies/tensordict-license.txt
License-File: licenses/dependencies/torch-license.txt
License-File: licenses/dependencies/torchvision-license.txt
Requires-Dist: torch>=2.6.0
Requires-Dist: torchvision>=0.5.0
Requires-Dist: tensordict>=0.7.0
Requires-Dist: numpy>=1.16.4
Requires-Dist: GitPython
Requires-Dist: onnx
Requires-Dist: onnxscript>=0.5.4
Dynamic: license-file

# RSL-RL

**RSL-RL** is a GPU-accelerated, lightweight learning library for robotics research. Its compact design allows
researchers to prototype and test new ideas without the overhead of modifying large, complex libraries. RSL-RL can also
be used out-of-the-box by installing it via [PyPI](https://pypi.org/project/rsl-rl-lib/), supports multi-GPU training,
and features common algorithms for robot learning.

## Key Features

- **Minimal, readable codebase** with clear extension points for rapid prototyping.
- **Robotics-first methods** including PPO and Student-Teacher Distillation.
- **High-throughput training** with native Multi-GPU support.
- **Proven performance** in numerous research publications.

## Learning Environments

RSL-RL is currently used by the following robot learning libraries:

- [Isaac Lab](https://github.com/isaac-sim/IsaacLab) (built on top of NVIDIA Isaac Sim)
- [Legged Gym](https://github.com/leggedrobotics/legged_gym) (built on top of NVIDIA Isaac Gym)
- [mjlab](https://github.com/mujocolab/mjlab) (built on top of MuJoCo Warp)
- [MuJoCo Playground](https://github.com/google-deepmind/mujoco_playground) (built on top of MuJoCo MJX and Warp)

## Installation

Before installing RSL-RL, ensure that Python `3.9+` is available. It is recommended to install the library in a virtual
environment (e.g. using `venv` or `conda`), which is often already created by the used environment library (e.g.
Isaac Lab). If so, make sure to activate it before installing RSL-RL.

### Installing RSL-RL as a dependency

```bash
pip install rsl-rl-lib
```

### Installing RSL-RL for development

```bash
git clone https://github.com/leggedrobotics/rsl_rl
cd rsl_rl
pip install -e .
```

## Citation

If you use RSL-RL in your research, please cite the [paper](https://arxiv.org/abs/2509.10771):

```text
@article{schwarke2025rslrl,
  title={RSL-RL: A Learning Library for Robotics Research},
  author={Schwarke, Clemens and Mittal, Mayank and Rudin, Nikita and Hoeller, David and Hutter, Marco},
  journal={arXiv preprint arXiv:2509.10771},
  year={2025}
}
```
