Verified project record

thuml/Time-Series-Library

time-series-library provides a neat code base for evaluating advanced deep time series models and developing new ones. It covers five mainstream tasks: long- and short-term forecasting, imputation, anomaly detection, and classification.

Project overview

By offering a single, organized code base that spans five mainstream time series tasks, time-series-library lowers the barrier to benchmarking existing models and prototyping new architectures.

Project type: Data Processing
Use cases: Data Analysis
Deployment: Refer to project documentation
License: MIT

Best for

Researchers who need a clean code base to evaluate advanced deep time series models or develop new ones.
Engineers working across multiple time series tasks such as forecasting, imputation, anomaly detection, and classification.

Key capabilities

Supports evaluating and developing models for long- and short-term time series forecasting.
Supports evaluating and developing models for time series imputation.
Supports evaluating and developing models for anomaly detection in time series.
Supports evaluating and developing models for time series classification.

Limitations and risks

Many of the library's benchmarks may no longer be meaningful for evaluating the effectiveness or progress of current research.
Due to limited bandwidth, maintainers will not be actively adding new features to this library.

Getting started

Install the updated Python requirements to run the library, or use the Docker deployment option as an alternative.

Evidence and sources

GitHub project description: A Library for Advanced Deep Time Series Models for General Time Series Analysis.
README: We provide a neat code base to evaluate advanced deep time series models or develop your model, which covers five mainstream tasks: long- and short-term forecasting, imputation, a…
README: Due to the limited bandwidth of the current maintainers, we will not be actively adding new features to this library. Since the library was originally released three years ago, ma…
README: we have newly added a [[zero-shot forecasting]](https://github.com/thuml/Time-Series-Library/blob/main/exp/exp_zero_shot_forecasting.py) feature in TSLib. You can try this script…
README: We have included [[TimeXer]](https://arxiv.org/abs/2402.19072), which defined a practical forecasting paradigm: Forecasting with Exogenous Variables.

Reviewed at: 2026-07-14T21:58:30Z