MBRL-Lib: A Modular Library for Model-based Reinforcement Learning

2021/04/20 by Pineda, Luis, Amos, Brandon, Zhang, Amy +2 · 3 citations
#Artificial Intelligence (cs.AI) #FOS: Computer and information sciences #FOS: Electrical engineering #Systems and Control (eess.SY) #electronic engineering #information engineering

paper · doi:10.48550/arxiv.2104.10159

Abstract

Model-based reinforcement learning is a compelling framework for data-efficient learning of agents that interact with the world. This family of algorithms has many subcomponents that need to be carefully selected and tuned. As a result the entry-bar for researchers to approach the field and to deploy it in real-world tasks can be daunting. In this paper, we present MBRL-Lib -- a machine learning library for model-based reinforcement learning in continuous state-action spaces based on PyTorch. MBRL-Lib is designed as a platform for both researchers, to easily develop, debug and compare new algorithms, and non-expert user, to lower the entry-bar of deploying state-of-the-art algorithms. MBRL-Lib is open-source at https://github.com/facebookresearch/mbrl-lib.

Cited by

• Double Horizon Model-Based Policy Optimization
• Learning Contact Dynamics through Touching: Action-conditional Graph Neural Networks for Robotic Peg Insertion
• Learning Causal Structure Distributions for Robust Planning

Related

• Baconian: A Unified Open-source Framework for Model-Based Reinforcement Learning
• Benchmarking Model-Based Reinforcement Learning
• Robust Model Based Reinforcement Learning Using L₁ Adaptive Control
• Intelligent Trainer for Model-Based Reinforcement Learning
• A Survey on Model-based Reinforcement Learning