The FaceMix Lora Model was a big challenge to create! Imagine mix two different characters faces in one head. I hope it can generate good images for you, but I am cannot sure that it will work fine with any Base models.
Try this prompt example and let me know your results:
"one head divided in two parts, half part of the head is an elephant , half part of the head is a porcupine , Pixar Style, high quality <lora:FaceMix:1>"
I tried it in A1111 with Turbo-Vision SuperFast SDXL model and it works very nice:
Steps: 8
CFG Scale: 2
Clip Drop: 2
Sampler: Euler A
SDXL Style: Adorable Kawaii / Astral Aura and others
512 x 512
Please post your images here that I can see your tries!