skrl: Modular and Flexible Library for Reinforcement Learning

Antonio Serrano-Muñoz; Dimitrios Chrysostomou; Simon Bøgh; Nestor Arana-Arexolaleiba

skrl: Modular and Flexible Library for Reinforcement Learning

Antonio Serrano-Muñoz, Dimitrios Chrysostomou, Simon Bøgh, Nestor Arana-Arexolaleiba

Research output: Contribution to journal › Journal article › Research › peer-review

108 Downloads (Pure)

Abstract

skrl is an open-source modular library for reinforcement learning written in Python and designed with a focus on readability, simplicity, and transparency of algorithm implementations. In addition to supporting environments that use the traditional interfaces from OpenAI Gym/Farama Gymnasium, DeepMind and others, it provides the facility to load, configure, and operate NVIDIA Isaac Gym, Isaac Orbit, and Omniverse Isaac Gym environments. Furthermore, it enables the simultaneous training of several agents with customizable scopes (subsets of environments among all available ones), which may or may not share resources, in the same run. The library's documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.

Original language	English
Journal	Journal of Machine Learning Research
Volume	24
Issue number	254
Pages (from-to)	1-9
Number of pages	9
ISSN	1533-7928
Publication status	Published - 15 Aug 2023

Bibliographical note

The library’s documentation can be found at https://skrl.readthedocs.io
and its source code is available on GitHub at https://github.com/Toni-SM/skrl.

Keywords

Reinforcement Learning (RL)
skrl
robot learning

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

Access to Document

JMLR 2023 - skrl_ModularandFlexibleLibraryforReinforcement LearningFinal published version, 316 KBLicence: CC BY 4.0

https://jmlr.org/papers/v24/23-0112.htmlLicence: CC BY 4.0

AUB Link

Search for the material in Aalborg University Library's search engine

R2P2: Networking for Research and Development of Human Interactive and Sensitive Robotics taking advantage of Additive Manufacturing
Chrysostomou, D. (PI), LI, C. (Project Participant), Arexolaleiba, N. A. (Project Participant) & Madsen, O. (Project Participant)
01/01/2020 → 31/12/2022
Project: Research

Cite this

@article{944a0737e28c47fdbdddd9686bfb79f5,

title = "skrl: Modular and Flexible Library for Reinforcement Learning",

abstract = "skrl is an open-source modular library for reinforcement learning written in Python and designed with a focus on readability, simplicity, and transparency of algorithm implementations. In addition to supporting environments that use the traditional interfaces from OpenAI Gym/Farama Gymnasium, DeepMind and others, it provides the facility to load, configure, and operate NVIDIA Isaac Gym, Isaac Orbit, and Omniverse Isaac Gym environments. Furthermore, it enables the simultaneous training of several agents with customizable scopes (subsets of environments among all available ones), which may or may not share resources, in the same run. The library's documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.",

keywords = "Reinforcement Learning (RL), skrl, robot learning",

author = "Antonio Serrano-Mu{\~n}oz and Dimitrios Chrysostomou and Simon B{\o}gh and Nestor Arana-Arexolaleiba",

note = "The library{\textquoteright}s documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.",

year = "2023",

month = aug,

day = "15",

language = "English",

volume = "24",

pages = "1--9",

journal = "Journal of Machine Learning Research",

issn = "1533-7928",

publisher = "MIT Press",

number = "254",

}

TY - JOUR

T1 - skrl: Modular and Flexible Library for Reinforcement Learning

AU - Serrano-Muñoz, Antonio

AU - Chrysostomou, Dimitrios

AU - Bøgh, Simon

AU - Arana-Arexolaleiba, Nestor

N1 - The library’s documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.

PY - 2023/8/15

Y1 - 2023/8/15

N2 - skrl is an open-source modular library for reinforcement learning written in Python and designed with a focus on readability, simplicity, and transparency of algorithm implementations. In addition to supporting environments that use the traditional interfaces from OpenAI Gym/Farama Gymnasium, DeepMind and others, it provides the facility to load, configure, and operate NVIDIA Isaac Gym, Isaac Orbit, and Omniverse Isaac Gym environments. Furthermore, it enables the simultaneous training of several agents with customizable scopes (subsets of environments among all available ones), which may or may not share resources, in the same run. The library's documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.

AB - skrl is an open-source modular library for reinforcement learning written in Python and designed with a focus on readability, simplicity, and transparency of algorithm implementations. In addition to supporting environments that use the traditional interfaces from OpenAI Gym/Farama Gymnasium, DeepMind and others, it provides the facility to load, configure, and operate NVIDIA Isaac Gym, Isaac Orbit, and Omniverse Isaac Gym environments. Furthermore, it enables the simultaneous training of several agents with customizable scopes (subsets of environments among all available ones), which may or may not share resources, in the same run. The library's documentation can be found at https://skrl.readthedocs.io and its source code is available on GitHub at https://github.com/Toni-SM/skrl.

KW - Reinforcement Learning (RL)

KW - skrl

KW - robot learning

M3 - Journal article

SN - 1533-7928

VL - 24

SP - 1

EP - 9

JO - Journal of Machine Learning Research

JF - Journal of Machine Learning Research

IS - 254

ER -

skrl: Modular and Flexible Library for Reinforcement Learning

Abstract

Bibliographical note

Keywords

UN SDGs

Access to Document

AUB Link

Fingerprint

Projects

R2P2: Networking for Research and Development of Human Interactive and Sensitive Robotics taking advantage of Additive Manufacturing

Cite this