This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHubThe SamplerSASolver node implements a custom sampling algorithm for diffusion models. It uses a predictor-corrector approach with configurable order settings and stochastic differential equation (SDE) parameters to generate samples from the input model.
Inputs
| Parameter | Data Type | Required | Range | Description |
|---|---|---|---|---|
model | MODEL | Yes | - | The diffusion model to use for sampling |
eta | FLOAT | Yes | 0.0 - 10.0 | Controls the step size scaling factor (default: 1.0) |
sde_start_percent | FLOAT | Yes | 0.0 - 1.0 | The starting percentage for SDE sampling (default: 0.2) |
sde_end_percent | FLOAT | Yes | 0.0 - 1.0 | The ending percentage for SDE sampling (default: 0.8) |
s_noise | FLOAT | Yes | 0.0 - 100.0 | Controls the amount of noise added during sampling (default: 1.0) |
predictor_order | INT | Yes | 1 - 6 | The order of the predictor component in the solver (default: 3) |
corrector_order | INT | Yes | 0 - 6 | The order of the corrector component in the solver (default: 4) |
use_pece | BOOLEAN | Yes | - | Enables or disables the PECE (Predict-Evaluate-Correct-Evaluate) method |
simple_order_2 | BOOLEAN | Yes | - | Enables or disables simplified second-order calculations |
Outputs
| Output Name | Data Type | Description |
|---|---|---|
sampler | SAMPLER | A configured sampler object that can be used with diffusion models |