The VPScheduler node is designed to generate a sequence of noise levels (sigmas) based on the Variance Preserving (VP) scheduling method.
The VPScheduler node is designed to generate a sequence of noise levels (sigmas) based on the Variance Preserving (VP) scheduling method. This sequence is crucial for guiding the denoising process in diffusion models, allowing for controlled generation of images or other data types.
Inputs
Parameter
Description
Data Type
steps
Specifies the number of steps in the diffusion process, affecting the granularity of the generated noise levels.
INT
beta_d
Determines the overall noise level distribution, influencing the variance of the generated noise levels.
FLOAT
beta_min
Sets the minimum boundary for the noise level, ensuring the noise does not fall below a certain threshold.
FLOAT
eps_s
Adjusts the starting epsilon value, fine-tuning the initial noise level in the diffusion process.
FLOAT
Outputs
Parameter
Description
Data Type
sigmas
A sequence of noise levels (sigmas) generated based on the VP scheduling method, used to guide the denoising process in diffusion models.
SIGMAS
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub