mapper/conf/model/image_encoder/dino.yaml

name: feature_extractor_DPT
backbone:
  pretrained: true
  frozen: true
  output_dim: ${model.latent_dim} # Match Latent Dimension