A clean and extensible PyTorch implementation of Masked Autoencoders Are Scalable Vision Learners

A PyTorch re-implementation of Mask Autoencoder training. SimpleMAE tries to be small, clean, interpretable and extensible.

pip install -r requirements.txt

(base) ➜ python main.py -c configs/imagenet.yaml --debug
Outputs will be saved to ../mae_out/imagenet-1116211657
Found 0 gpu(s)
Param num: 76445440
Epoch 0 Train loss 1.5653417110443115
Output has been saved to ../mae_out/imagenet-1116211657