skrl

Modular Reinforcement Learning (RL) library (implemented in PyTorch, JAX, and NVIDIA Warp) with support for Gymnasium/Gym, NVIDIA Isaac Lab, MuJoCo Playground a

What is skrl?

Modular Reinforcement Learning (RL) library (implemented in PyTorch, JAX, and NVIDIA Warp) with support for Gymnasium/Gym, NVIDIA Isaac Lab, MuJoCo Playground and other environments

Overview of skrl

What it is

skrl is an open-source modular library for Reinforcement Learning written in Python. It is implemented in PyTorch, JAX, and NVIDIA Warp. The library is designed with a focus on modularity, readability, simplicity, and transparency of algorithm implementation. It supports OpenAI Gym, Farama Gymnasium and PettingZoo, ManiSkill, and other environment interfaces. It also allows loading and configuring NVIDIA Isaac Lab and MuJoCo Playground environments. The library enables agents' simultaneous training by scopes (subsets of environments among all available environments), which may or may not share resources, in the same run.

Key features

  • Modular design with implementations in PyTorch, JAX, and NVIDIA Warp.
  • Support for multiple environment interfaces: OpenAI Gym, Farama Gymnasium, PettingZoo, ManiSkill, NVIDIA Isaac Lab, MuJoCo Playground, and others.
  • Ability to load and configure NVIDIA Isaac Lab and MuJoCo Playground environments.
  • Simultaneous training of agents by scopes (subsets of environments) in the same run.
  • Focus on readability, simplicity, and transparency of algorithm implementation.

Use cases

skrl can be used for research and development in reinforcement learning, particularly when modularity and transparency are important. It is suitable for training agents in a variety of simulated environments, including those from NVIDIA Isaac Lab and MuJoCo Playground. The library's support for multiple backends (PyTorch, JAX, NVIDIA Warp) allows users to choose the framework that best fits their needs. The ability to train agents simultaneously on subsets of environments is useful for experiments requiring environment partitioning.

Getting started

See the repository README for installation and quick-start instructions.

License and pricing

The library is licensed under the MIT license. It is open source and free to use.

Based on the project's public README and repository metadata.

Add this badge to your README

Featured on YunoDB Badge styles & help

Maintain this project? Show that it is featured on YunoDB.

Markdown[![Featured on YunoDB](https://yunodb.com/badge/toni-smskrl.svg)](https://yunodb.com/tools/toni-smskrl?utm_source=github&utm_medium=badge)
HTML<a href="https://yunodb.com/tools/toni-smskrl?utm_source=github&utm_medium=badge"><img src="https://yunodb.com/badge/toni-smskrl.svg" alt="Featured on YunoDB" height="20"></a>