config.json

{
  "_class_name": "HeadSegmentationModel",
  "_diffusers_version": "0.30.0",
  "encoder_depth": 5,
  "in_channels": 3,
  "input_resolution": 512,
  "out_channels": 2
}