median-research-group/LibMTL
A PyTorch Library for Multi-Task Learning
Provides modular implementations of 16+ gradient-based optimization strategies (GradNorm, MGDA, Nash-MTL, FAMO, etc.) and 8 architectural patterns for balancing competing task objectives. Features a unified benchmark framework with standardized datasets, metrics, and evaluation protocols enabling reproducible cross-algorithm comparisons. Designed with extensible components for rapidly prototyping novel MTL methods or adapting existing approaches to new domains.
2,531 stars and 87 monthly downloads. No commits in the last 6 months. Available on PyPI.
Stars
2,531
Forks
232
Language
Python
License
MIT
Category
Last pushed
May 14, 2025
Monthly downloads
87
Commits (30d)
0
Dependencies
3
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/median-research-group/LibMTL"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Related frameworks
qdrant/quaterion
Blazing fast framework for fine-tuning similarity learning models
zeiss-microscopy/OAD
Collection of tools and scripts useful to automate microscopy workflows in ZEN Blue using Python...
lucidrains/PoPE-pytorch
Efficient implementation (and explorations) into polar coordinate positional embedding (PoPE) -...
Zehong-Wang/Awesome-Weight-Space-Learning
A collection of weight space learning including papers, codes, and datasets.
snimu/rebasin
Apply methods described in "Git Re-basin"-paper [1] to arbitrary models --- [1] Ainsworth et...