[Feature Request]: Sigma Churn has different effects greater than 1

[NouberNou]

Is there an existing issue for this?

• I have searched the existing issues and checked the recent builds/commits

What would your feature do ?

Sigma Churn for Euler, Heun, and DPM2 samplers have different (positive) effects when set greater than 1 up to 20 (where it then begins to stagnate in effect). This can be tested using the X/Y/Z grid.

I would suggest setting the maximum value of the s_churn option for these samplers to 20 instead of 1 in modules/shared.py.

Proposed workflow

Adding the s_churn option to quick options should allow a maximum value of 20.

Additional information

I actually can't find any documentation on why values greater than 1 work, but they do, and they often provide significantly more intricate detail on high-step inferences. Possibly it is wrapping and I am just missing these same churns 0..1 someplace but all I know is values greater than 1 significantly differ than most 0-1 values.

[catboxanon]

Resolved in #12356