search

mit-han-lab / efficientvit

EfficientViT is a new family of vision models for efficient high-resolution vision.
Apache License 2.0
1.59k stars 141 forks source link

train the image encoder of EfficientViT-SAM #102

Open Yangenhui98 opened 2 months ago

Yangenhui98 commented 2 months ago

How to train the image encoder of EfficientViT-SAM using SAM’s image encoder as the teacher，where is the code located？