Adapting EfficientViT models for use with MMSegmentation.
Note: Some model variants had backbone configurations listed, but did not have semantic segmentation head configurations listed in the paper or official repo and are thus not implemented here:
- EfficientViT-L0
- EfficientViT-L3