-
You should download MAE pre-trained weights from the official repo of the paper Masked Autoencoders are scalable vision learners before start training(fine-tuning) with our IML-ViT.
-
By default, our commend for training takes this dir as the place where pre-trained weights are. Therefore, you should place your downloaded weights in this directory, or re-direct to your custom dir of MAE weights.