This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHubThe SamplerDPMAdaptative node implements an adaptive DPM (Diffusion Probabilistic Model) sampler that automatically adjusts step sizes during the sampling process. It uses tolerance-based error control to determine optimal step sizes, balancing computational efficiency with sampling accuracy. This adaptive approach helps maintain quality while potentially reducing the number of steps needed.
Inputs
| Parameter | Data Type | Required | Range | Description |
|---|---|---|---|---|
order | INT | Yes | 2-3 | The order of the sampler method (default: 3) |
rtol | FLOAT | Yes | 0.0-100.0 | Relative tolerance for error control (default: 0.05) |
atol | FLOAT | Yes | 0.0-100.0 | Absolute tolerance for error control (default: 0.0078) |
h_init | FLOAT | Yes | 0.0-100.0 | Initial step size (default: 0.05) |
pcoeff | FLOAT | Yes | 0.0-100.0 | Proportional coefficient for step size control (default: 0.0) |
icoeff | FLOAT | Yes | 0.0-100.0 | Integral coefficient for step size control (default: 1.0) |
dcoeff | FLOAT | Yes | 0.0-100.0 | Derivative coefficient for step size control (default: 0.0) |
accept_safety | FLOAT | Yes | 0.0-100.0 | Safety factor for step acceptance (default: 0.81) |
eta | FLOAT | Yes | 0.0-100.0 | Stochasticity parameter (default: 0.0) |
s_noise | FLOAT | Yes | 0.0-100.0 | Noise scaling factor (default: 1.0) |
Outputs
| Output Name | Data Type | Description |
|---|---|---|
sampler | SAMPLER | Returns a configured DPM adaptive sampler instance |