Documentation Index
Fetch the complete documentation index at: /llms.txt
Use this file to discover all available pages before exploring further.
This node provides an LCM (Latent Consistency Model) sampler with adjustable per-step noise.
s_noise acts as a multiplier on the model’s training noise scale, and the noise level can vary from the first step to the last step. The configured sampler can then be plugged into a sampling workflow.
| Parameter | Description | Data Type | Required | Range |
|---|---|---|---|---|
s_noise | Per-step noise multiplier at the first step (1.0 = match training). Default: 1.0. | FLOAT | Yes | 0.0 to 64.0 (step: 0.01) |
s_noise_end | Per-step noise multiplier at the last step. Set equal to s_noise for a constant schedule. Default: 1.0. | FLOAT | Yes | 0.0 to 64.0 (step: 0.01) |
noise_clip_std | Clamp per-step noise to +/- N*std. 0 disables. Default: 0.0. | FLOAT | Yes | 0.0 to 10.0 (step: 0.01) |
| Output Name | Description | Data Type |
|---|---|---|
SAMPLER | The configured LCM sampler object, ready to be used in a sampling workflow. | SAMPLER |
This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub
0d18f2f977ddadeedcd7807233b48ebcc4e94c6213f8540b9037a45a9c70c6cfWas this page helpful?