Type: huggingface.image_to_image.VAEEncode

Namespace: huggingface.image_to_image

Description

Encodes an image into latents using a VAE. image -> tensor (TorchTensor)

Properties

Property Type Description Default
model hf.vae The VAE model to use. {'type': 'hf.vae', 'repo_id': '', 'path': None, 'variant': None, 'allow_patterns': None, 'ignore_patterns': None}
image image Input image to encode. {'type': 'image', 'uri': '', 'asset_id': None, 'data': None}
scale_factor float Scaling factor applied to latents (e.g., 0.18215 for SD15) 0.18215

Outputs

Output Type Description
output torch_tensor  

Metadata

Browse other nodes in the huggingface.image_to_image namespace.