facebookresearch/mae
A PyTorch implementation of Masked Autoencoders (MAE), a self-supervised vision transformer architecture for scalable image representation learning.
Not currently ranked — collecting fresh signals.
star history
This repository provides a PyTorch re-implementation of the MAE paper on masked autoencoders for vision, originally released in TensorFlow+TPU. It includes pre-training code for self-supervised learning on images, fine-tuning scripts with pre-trained ViT checkpoints across Base/Large/Huge sizes, and an interactive visualization demo for exploring learned representations. The implementation builds on the timm library and the DeiT repository.
Frequently asked
- What is facebookresearch/mae?
- A PyTorch implementation of Masked Autoencoders (MAE), a self-supervised vision transformer architecture for scalable image representation learning.
- Is mae open source?
- Yes — facebookresearch/mae is an open-source project tracked on heatdrop.
- What language is mae written in?
- facebookresearch/mae is primarily written in Python.
- How popular is mae?
- facebookresearch/mae has 8.4k stars on GitHub.
- Where can I find mae?
- facebookresearch/mae is on GitHub at https://github.com/facebookresearch/mae.