ComfyUI Node
pi-Flow Sampler
Uses the provided model, positive conditioning to denoise the latent image.
pi-Flow Sampler
- model
- conditioning
- latent_image
- LATENT
◄seed0►
◄steps4►
◄substeps128►
◄final_step_size_scale0.50►
◄diffusion_coefficient0.00►
◄gm_temperatureauto►
◄manual_gm_temperature1.00►
◄denoise1.00►
CategoryLakonLab
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | The model used for denoising the input latent. | |
| seed | INT | 00–18446744073709550000 | The random seed used for creating the noise. |
| steps | INT | 41–10000 | The number of network steps used in the denoising process. |
| substeps | INT | 1281–10000 | The number of policy sub-steps used in the denoising process. |
| final_step_size_scale | FLOAT | 0.500–1 | The size of the final step relative to other steps. |
| diffusion_coefficient | FLOAT | 0.000–1000000 | The coefficient controlling the stochasticity of the sampling process. 0.0 is deterministic. 1.0 is standard DDPM stochasticity. |
| gm_temperature | COMBO | auto | The GMFlow temperature setting. |
| manual_gm_temperature | FLOAT | 1.000–1 | The GMFlow temperature to use if gm_temperature is set to manual. |
| conditioning | CONDITIONING | The conditioning describing the attributes you want to include in the image. | |
| latent_image | LATENT | The latent image to denoise. | |
| denoise | FLOAT | 1.000–1 | The amount of denoising applied, lower values will maintain the structure of the initial image allowing for image to image sampling. |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| LATENT | LATENT | The denoised latent. |