{
  "_class_name": "VersatileDiffusionPipeline",
  "_diffusers_version": "0.8.0.dev0",
  "image_encoder": [
    "transformers",
    "CLIPVisionModelWithProjection"
  ],
  "image_feature_extractor": [
    "transformers",
    "CLIPImageProcessor"
  ],
  "image_unet": [
    "diffusers",
    "UNet2DConditionModel"
  ],
  "scheduler": [
    "diffusers",
    "DDIMScheduler"
  ],
  "text_encoder": [
    "transformers",
    "CLIPTextModelWithProjection"
  ],
  "text_unet": [
    "versatile_diffusion",
    "UNetFlatConditionModel"
  ],
  "tokenizer": [
    "transformers",
    "CLIPTokenizer"
  ],
  "vae": [
    "diffusers",
    "AutoencoderKL"
  ]
}