LibMTL: A Python Library for Multi-Task Learning
This library addresses the need for a comprehensive and easy-to-use tool for researchers and practitioners working on MTL, enabling faster development and fair comparison of methods, though it is incremental as it builds on existing MTL approaches.
The paper introduces LibMTL, an open-source Python library for Multi-Task Learning (MTL) that provides a unified framework supporting 12 loss weighting strategies, 7 architectures, and 84 combinations to facilitate reproducible and extensible MTL research and applications.
This paper presents LibMTL, an open-source Python library built on PyTorch, which provides a unified, comprehensive, reproducible, and extensible implementation framework for Multi-Task Learning (MTL). LibMTL considers different settings and approaches in MTL, and it supports a large number of state-of-the-art MTL methods, including 12 loss weighting strategies, 7 architectures, and 84 combinations of different architectures and loss weighting methods. Moreover, the modular design in LibMTL makes it easy-to-use and well extensible, thus users can easily and fast develop new MTL methods, compare with existing MTL methods fairly, or apply MTL algorithms to real-world applications with the support of LibMTL. The source code and detailed documentations of LibMTL are available at https://github.com/median-research-group/LibMTL and https://libmtl.readthedocs.io, respectively.