EasyUse/Loadersgenerated

EasyLoader (Zero123)(zero123Loader)

The EasyLoader (Zero123) node (zero123Loader) is an EasyUse loader for Zero123-style checkpoint models.

zero123 Loader

init_image
pipe
model
vae
ckpt_name
COMBO
vae_name
COMBO
empty_latent_width
INT
empty_latent_height
INT
batch_size
1
elevation
0
azimuth
0
Easy Use

Description

The EasyLoader (Zero123) node (zero123Loader) is an EasyUse loader for Zero123-style checkpoint models. It loads a checkpoint and VAE, conditions them on an input image, and constructs an empty latent with configurable dimensions and camera parameters. It is intended for novel-view synthesis workflows where a single input image is used to generate additional views with controlled elevation and azimuth.

Inputs

  • ckpt_name: The checkpoint to load. Select from the available Zero123-style checkpoint files.
  • vae_name: The VAE to load. Select from the available VAE files.
  • init_image: The input image used as the conditioning view for the Zero123 model.
  • empty_latent_width: The width of the empty latent batch to create.
  • empty_latent_height: The height of the empty latent batch to create.
  • batch_size: Number of latents to generate. Default is 1; allowed range is 1 to 64.
  • elevation: Elevation angle in degrees. Default is 0; allowed range is -180 to 180.
  • azimuth: Azimuth angle in degrees. Default is 0; allowed range is -180 to 180.

Outputs

  • PIPE_LINE: An EasyUse pipeline object containing the loaded model, VAE, conditioning, and latent for use with other EasyUse nodes.
  • MODEL: The loaded Zero123 model.
  • VAE: The loaded VAE.

Usage Notes

Use the elevation and azimuth inputs to control the relative camera angle of the generated view. The init_image provides the visual reference, while the empty latent dimensions determine the spatial size of the generated output. The batch_size input is useful for generating multiple views or samples at once.

Comments

Sign in with GitHub to join the discussion.

Loading comments…