ComfyUI Node: Noise Inversion

Authored by shiimizu

Created 10 months ago

Updated 2 months ago

296 stars

Category

sampling

Inputs

model MODEL

positive CONDITIONING

negative CONDITIONING

latent_image LATENT

image IMAGE

steps INT

retouch FLOAT

renoise_strength FLOAT

renoise_kernel_size INT

Outputs

LATENT

Extension: Tiled Diffusion & VAE for ComfyUI

The extension enables large image drawing & upscaling with limited VRAM via the following techniques: 1.Two SOTA diffusion tiling algorithms: a/Mixture of Diffusers and a/MultiDiffusion 2.pkuliyi2015's Tiled VAE algorithm.

Authored by shiimizu