Visit ComfyUI Online for ready-to-use ComfyUI environment
Normalize conditioning data mean before CFG for improved AI image quality.
The Pre CFG subtract mean node is designed to enhance the quality of your AI-generated images by normalizing the mean of the conditioning data before the classifier-free guidance (CFG) process. This node ensures that the mean value of the conditioning data is subtracted, which can help in reducing biases and improving the overall consistency of the generated images. By applying this normalization step, the node aims to produce more balanced and visually appealing results, making it a valuable tool for AI artists looking to refine their outputs.
This parameter specifies the model to which the mean subtraction will be applied. It is essential for defining the context in which the node operates, ensuring that the correct model is used for the normalization process.
This parameter determines the sigma value at which the mean subtraction process starts. It is a floating-point value with a default of 15.0, a minimum of 0.0, and a maximum of 1000.0. Adjusting this value allows you to control the point in the denoising process where the mean subtraction begins, which can impact the final image quality.
This parameter sets the sigma value at which the mean subtraction process ends. It is a floating-point value with a default of 1.0, a minimum of 0.0, and a maximum of 1000.0. By configuring this value, you can define the range within which the mean subtraction is active, providing finer control over the normalization process.
This boolean parameter enables or disables the mean subtraction process. By default, it is set to True. When enabled, the node performs the mean subtraction; when disabled, the node bypasses this step, allowing you to easily toggle the effect on and off.
The output parameter is the model with the mean subtraction applied. This model has undergone the normalization process, resulting in potentially improved image quality and consistency. The output model can then be used in subsequent steps of your AI art generation workflow.
start_at_sigma
and end_at_sigma
values to fine-tune the range within which the mean subtraction is applied. This can help you achieve the desired balance and consistency in your generated images.enabled
parameter to quickly compare the effects of the mean subtraction by toggling it on and off. This can help you determine whether the normalization process is beneficial for your specific use case.start_at_sigma
or end_at_sigma
values are out of the acceptable range.enabled
parameter is set to False, so the mean subtraction process is bypassed.enabled
parameter to True to activate the mean subtraction process.© Copyright 2024 RunComfy. All Rights Reserved.