An experiment combining Asanagi with InCase. Please support both artists as these Loras were likely trained without their permission and they are both fantastic.
Model: waiANIHENTAIPONYXL
Loras: 1nc4s3XLP, asanagi lora, enhancerV4-pony
You can extract the tags from the images, they changed a lot.
Please support both artists if you like this style. Without them these gens would not be possible.
Experiment was basically just to create a girl and try to mesh both artist’s styles. I love Asanagi’s expressions, they are beautiful. I love InCase’s bodies. BEST OF BOTH WORLDS ACHIEVED.
Fascinatingly, the AI tended to really favor InCase’s style, even with lower weights. I didn’t keep the weight for Asanagi’s very high admittedly, I was trying to avoid too much of them bleeding in. Just the psychotic expressions.
This is one instance where ADetail actually really improved images. You can see in some generations where I turned it off to see what would happen. Generally, faces and hands were a lot better. Generally.
For some reason it just wanted her wearing a white outfit. Maybe I was temporarily dyslexic and didn’t see “white dress”, but it was really into it.
ANYWAY. Enjoy the gallery.
It seems like WordPress strips metadata so here’s an archive in case you want to view the prompt.
