# thuml/Time-Series-Library

A Library for Advanced Deep Time Series Models for General Time Series Analysis.

Repository: https://github.com/thuml/Time-Series-Library
Canonical: https://ross.abutalabs.com/products/time-series-library
Language: Python
License: MIT
License Family: permissive
Topics: deep-learning, time-series, time-series-analysis
Last push: 2026-04-18T21:01:40+00:00

## Health v2 (maintenance only)
Score: 66/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 78, release rhythm 35, longevity 92
- inputs: {"age_days": 1297, "days_push": 137, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 12785, forks 1986 (observed 2026-08-28T04:10:59.761616+00:00)

## What it is
TSLib is an open-source Python library providing a unified codebase of advanced deep learning models for general time series analysis. It supports long- and short-term forecasting, imputation, anomaly detection, and classification, and is widely used as a research benchmark.

## Use cases
- benchmark deep time series forecasting models
- run anomaly detection on time series data
- impute missing values in time series
- classify time series with deep learning
- develop and evaluate a new time series model
- zero-shot forecasting with large time series models

## When to choose
- you are a researcher evaluating or developing deep time series models
- you need consistent baselines across forecasting, imputation, anomaly detection, and classification tasks
- you want a neat, unified codebase for time series experiments

## When to avoid
- you need production-ready time series pipelines rather than research code
- you require actively developed new features, since the project is no longer actively adding them
- you need up-to-date benchmarks, as many original ones may be saturated

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning, benchmarking
- domain: time-series, deep-learning, machine-learning, data-science
- platform: python
- tags: time-series-forecasting, anomaly-detection, imputation, classification, research-benchmark, deep-time-series-models

## Member repositories
- thuml/Time-Series-Library (main) score 66

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:10:59.761616+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T17:13:47.900813+00:00, confidence not recorded.
  - readme: https://github.com/thuml/Time-Series-Library (fetched 2026-08-28T04:10:59.761616+00:00, sha e774e2ba3b92)
- Data as of 2026-08-30T08:39:29.467469+00:00.