search

mit-han-lab / efficientvit

EfficientViT is a new family of vision models for efficient high-resolution vision.
Apache License 2.0
1.6k stars 142 forks source link

Any plan for dinov2? #54

Open twmht opened 7 months ago

twmht commented 7 months ago

DinoV2 has shown remarkable performance on downstream tasks, but its use of Vision Transformer (ViT) is computationally inefficient. Do you have plans to train an efficientVIT version of DinoV2?

han-cai commented 7 months ago

Hi twmht,

Thank you for your interest. We do have this in our plan. If you want to get informed about our updates, please join our mailing list here or star/watch this repo.

Best, Han

JayKarhade commented 6 months ago

This is awesome! What would be a tentative release date for dinov2?