Skip to main content
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
The FreeU_V2 node applies a frequency-based enhancement to diffusion models by modifying the U-Net architecture. It scales different feature channels using configurable parameters to improve image generation quality without requiring additional training. The node works by patching the model’s output blocks to apply scaling factors to specific channel dimensions.

Inputs

ParameterData TypeRequiredRangeDescription
modelMODELYes-The diffusion model to apply FreeU enhancement to
b1FLOATYes0.0 - 10.0Backbone feature scaling factor for the first block (default: 1.3)
b2FLOATYes0.0 - 10.0Backbone feature scaling factor for the second block (default: 1.4)
s1FLOATYes0.0 - 10.0Skip feature scaling factor for the first block (default: 0.9)
s2FLOATYes0.0 - 10.0Skip feature scaling factor for the second block (default: 0.2)

Outputs

Output NameData TypeDescription
modelMODELThe enhanced diffusion model with FreeU modifications applied