Does SAS offer a proc step to fit nonparametric mixtures of regression models?

[DaLack]

I have estimated a finite mixture regression model using proc fmm. But I am more interested in a nonparametric version, explained for example here:

https://methodology.psu.edu/media/techreports/09-93.pdf

Is there a possibilty to relax the linear assumption using proc fmm or does an adequate proc step exist?

Thanks in advance.