{
  "model_type": "stable-diffusion",
  "architectures": ["StableDiffusion"],
  "tokenizer_class": "CLIPTextModel"
}