thuml/Time-Series-Library
A Library for Advanced Deep Time Series Models for General Time Series Analysis.
Supports five mainstream time series tasks (forecasting with exogenous variables, imputation, anomaly detection, classification) with unified benchmarking across 30+ state-of-the-art models including Transformers, state-space models (Mamba), and decomposition-based architectures. Features zero-shot evaluation capabilities for large language model-based time series forecasting and standardized datasets with rigorous experimental protocols to address benchmark saturation issues.
11,714 stars. Actively maintained with 7 commits in the last 30 days.
Stars
11,714
Forks
1,868
Language
Python
License
MIT
Category
Last pushed
Feb 23, 2026
Commits (30d)
7
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/thuml/Time-Series-Library"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Compare
Related frameworks
sktime/sktime
A unified framework for machine learning with time series
aeon-toolkit/aeon
A toolkit for time series machine learning and deep learning
skforecast/skforecast
Time series forecasting with machine learning models
Nixtla/neuralforecast
Scalable and user friendly neural :brain: forecasting algorithms.
Nixtla/mlforecast
Scalable machine 🤖 learning for time series forecasting.