Skip to main content
The ModelSamplingContinuousV node modifies a model’s sampling behavior by applying continuous V-prediction sampling parameters. It creates a clone of the input model and configures it with custom sigma range settings for advanced sampling control. This allows users to fine-tune the sampling process with specific minimum and maximum sigma values.

Inputs

ParameterDescriptionData TypeRequiredRange
modelThe input model to be modified with continuous V-prediction samplingMODELYes-
samplingThe sampling method to apply (currently only V-prediction is supported)STRINGYes"v_prediction"
sigma_maxThe maximum sigma value for sampling (default: 500.0)FLOATYes0.0 - 1000.0
sigma_minThe minimum sigma value for sampling (default: 0.03)FLOATYes0.0 - 1000.0

Outputs

Output NameDescriptionData Type
modelThe modified model with continuous V-prediction sampling appliedMODEL
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 8095b5024c0d33011f6a81ed496cf1711981701e0f35f9527646b150f5033d45