# zhouhaoyi/Informer2020

The GitHub repository for the paper "Informer" accepted by AAAI 2021.

Repository: https://github.com/zhouhaoyi/Informer2020
Canonical: https://ross.abutalabs.com/products/informer2020
Language: Python
License: Apache-2.0
License Family: permissive
Topics: deep-learning, self-attention, time-series, pytorch, forecasting, transformer
Last push: 2025-06-20T06:42:54+00:00

## Health v2 (maintenance only)
Score: 44/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 27, release rhythm 35, longevity 100
- inputs: {"age_days": 2095, "days_push": 439, "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 6516, forks 1297 (observed 2026-08-28T04:09:44.854460+00:00)

## What it is
The official PyTorch implementation of Informer, an efficient Transformer architecture for long sequence time-series forecasting that won the AAAI 2021 Best Paper award. It introduces ProbSparse attention to select active queries and enable efficient forecasting on very long time series.

## Use cases
- forecast long time series with transformers
- implement ProbSparse attention in PyTorch
- reproduce Informer paper experiments
- benchmark long sequence forecasting models
- predict multivariate time series like electricity or traffic
- learn efficient transformer architectures for time series

## When to choose
- you need to forecast long time series sequences with a transformer model
- you want to reproduce or build on the Informer AAAI'21 paper
- you need a PyTorch baseline for long sequence time-series forecasting research

## When to avoid
- you need a production-ready forecasting service rather than research code
- you want actively developed features or broad model support beyond Informer
- you need a general-purpose time series library with many model options

## Facets
- artifact type: library
- maturity: maintenance
- function: deep-learning, machine-learning
- domain: deep-learning, time-series, machine-learning
- platform: python
- tags: transformer, time-series-forecasting, self-attention, probsparse-attention, pytorch, research-code, aaai-2021, gpu

## Member repositories
- zhouhaoyi/Informer2020 (main) score 44

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:44.854460+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 2026-08-29T17:44:26.454160+00:00, confidence not recorded.
  - readme: https://github.com/zhouhaoyi/Informer2020 (fetched 2026-08-28T04:09:44.854460+00:00, sha b003e318136b)
- Data as of 2026-08-30T08:39:29.467469+00:00.