mirror of
https://github.com/hanzoui/workflow-templates-broken-zap-2026-06-08.git
synced 2026-10-30 20:44:35 +00:00
1.5 KiB
1.5 KiB
InstructPixToPixConditioning
Category: conditioning/instructpix2pix
Display Name: positive
Description
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
The InstructPixToPixConditioning node prepares conditioning data for InstructPix2Pix image editing by combining positive and negative text prompts with image data. It processes input images through a VAE encoder to create latent representations and attaches these latents to both positive and negative conditioning data. The node automatically handles image dimensions by cropping to multiples of 8 pixels for compatibility with the VAE encoding process.
Inputs
| Parameter | Type | Default | Description |
|---|---|---|---|
| positive | Conditioning | — | |
| negative | Conditioning | — | |
| vae | Vae | — | |
| pixels | Image | — |
Outputs
| Output Name | Data Type | Description |
|---|---|---|
positive |
CONDITIONING | Positive conditioning data with attached latent image representation |
negative |
CONDITIONING | Negative conditioning data with attached latent image representation |
latent |
LATENT | Empty latent tensor with the same dimensions as the encoded image |
Source: comfy_extras/nodes_ip2p.py
Used in 4 template(s)