Allow skipping weight initialisation for Masked Vision Transformers.

lightly-ai / lightly

A python library for self-supervised learning on images.

https://docs.lightly.ai/self-supervised-learning/

MIT License

3.09k stars 264 forks source link

Allow skipping weight initialisation for Masked Vision Transformers. #1511

Closed ersi-lightly closed 1 month ago

ersi-lightly commented 7 months ago

Allow skipping weight initializsation for MaskedVisionTransformerTIMM, MaskedVisionTransformerTorchvision and MAEDecoderTIMM. This will enable to load a pretrained vit model.