# microsoft/SimMIM

This is an official implementation for "SimMIM: A Simple Framework for Masked Image Modeling".

Repository: https://github.com/microsoft/SimMIM
Canonical: https://ross.abutalabs.com/products/simmim
Homepage: https://arxiv.org/abs/2111.09886
Language: Python
License: MIT
License Family: permissive
Topics: self-supervised-learning, masked-image-modeling, image-classification, swin-transformer
Last push: 2022-09-29T15:17:40+00:00

## Health v2 (maintenance only)
Score: 32/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 0, release rhythm 35, longevity 100
- inputs: {"age_days": 1749, "days_push": 1434, "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 1048, forks 107 (observed 2026-08-28T04:03:22.428781+00:00)

## What it is
Official PyTorch implementation of SimMIM, a simple framework for masked image modeling (self-supervised visual pre-training) from Microsoft Research. It provides pre-training and fine-tuning code plus pretrained checkpoints for Swin Transformer and ViT models on ImageNet-1K.

## Use cases
- pretrain vision transformers with masked image modeling
- self-supervised pretraining on ImageNet
- fine-tune Swin Transformer for image classification
- reproduce SimMIM CVPR 2022 paper results
- pretrain large vision models with limited data
- compare masked image modeling approaches

## When to choose
- you need self-supervised pretraining for Swin or ViT backbones
- you want to reproduce or build on the SimMIM paper
- you need pretrained checkpoints for image classification research

## When to avoid
- you need general-purpose production computer vision tools
- you want actively developed features (code merged into Swin-Transformer repo)
- you need non-PyTorch or non-vision ML frameworks

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, image-processing
- domain: computer-vision, deep-learning, machine-learning
- platform: python
- tags: masked-image-modeling, self-supervised-learning, vision-transformer, pretraining, image-classification, swin-transformer, research-code, gpu

## Member repositories
- microsoft/SimMIM (main) score 32

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:03:22.428781+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 2026-08-30T07:00:45.616887+00:00, confidence not recorded.
  - readme: https://github.com/microsoft/SimMIM (fetched 2026-08-28T04:03:22.428781+00:00, sha 3349610e6a25)
  - homepage: https://arxiv.org/abs/2111.09886 (fetched 2026-08-29T13:02:40.782053+00:00, sha f346c7824189)
  - site_page: https://info.arxiv.org/about/donate.html (fetched 2026-08-29T13:02:40.784794+00:00, sha cca9c3a11c56)
  - site_page: https://info.arxiv.org/about/ourmembers.html (fetched 2026-08-29T13:02:40.788446+00:00, sha 47cbc55ff1de)
  - site_page: https://info.arxiv.org/about (fetched 2026-08-29T13:02:40.792086+00:00, sha a1f16f915a9a)
  - site_page: https://info.arxiv.org/labs/index.html (fetched 2026-08-29T13:02:40.786592+00:00, sha b14a8d05a0ec)
- Data as of 2026-08-30T08:39:29.467469+00:00.