How to use vikp/surya_layout4 with Transformers:
# Load model directly from transformers import EfficientViTForSemanticSegmentation model = EfficientViTForSemanticSegmentation.from_pretrained("vikp/surya_layout4", dtype="auto")
