# facebookresearch/deit

Official DeiT repository

Repository: https://github.com/facebookresearch/deit
Canonical: https://ross.abutalabs.com/products/deit
Language: Python
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2024-03-15T13:20:37+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 2079, "days_push": 901, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 4355, forks 592 (observed 2026-08-28T04:08:46.436632+00:00)

## What it is
Official PyTorch repository for DeiT and related vision transformer architectures (CaiT, ResMLP, PatchConvnet, DeiT III), providing training, evaluation code and pretrained models for data-efficient image classification. It accompanies published research papers from Facebook AI Research.

## Use cases
- train vision transformers with limited data
- download pretrained image classification models
- reproduce DeiT research results
- fine-tune image transformers on custom datasets
- compare vision transformer architectures like CaiT and ResMLP
- learn knowledge distillation through attention

## When to choose
- you need data-efficient vision transformer training or pretrained checkpoints
- you want to reproduce or build on DeiT-family research papers
- you need PyTorch implementations of ViT variants for image classification

## When to avoid
- you need a production image classification service rather than research code
- you work outside PyTorch
- you need general-purpose computer vision beyond image classification

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: computer-vision, deep-learning, image-processing
- platform: python
- tags: vision-transformer, pytorch, image-classification, knowledge-distillation, research-code, pretrained-models, gpu

## Member repositories
- facebookresearch/deit (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:46.436632+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 2026-08-29T18:21:30.401350+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/deit (fetched 2026-08-28T04:08:46.436632+00:00, sha 7f7651dcb780)
- Data as of 2026-08-30T08:39:29.467469+00:00.