try any of these for default look or put in Neg to reduce chance (making changes)
may need to add 3d, or realistic, in neg for 2d/anime styles or lowered weight
ponyXL / iLLXL - TW is Alcinaxl,
iLL - v2 may be more accurate but may have more 3d influence
ILL - V3 have more in-game face and expression training
face/body
Alcina, yellow eyes, or grey eyes, black hair, medium hair, makeup
adding pale skin, may not be needed
can try, lipstick, mature female, make up, veiny skin, curvy,
Outfit
white dress, hat, single earring, cleavage, black gloves, pantyhose,
pearl necklace, or silver necklace,
dress have a flower brooch, or black rose,
did not train black claws, or at least did not focus them in training, add long,
hat, black hat, large hat, or sun hat, etc,
Description
FAQ
Comments (8)
I just got around to looking for it in your pile of Loras today! Are you reading my mind?))
well if anyone set my loras to most downloaded or highest rated this was number 2 I think. Im playing replaying RE8 atm to update the other character loras
@True_Might No, really. You're probably a mind reader!
And will you make her daughters? Or sisters? I don't remember what they are.
By the way, I'm currently creating a video with this LADY! It looks absolutely stunning! Eyes will drool with aesthetic pleasure!
@DaviDevil yeah Im planning to do the daughters, the three of them would be new loras for me. it takes me about 2-3 hours to train per lora. so they may be done last after I update the others.
@True_Might I'm really looking forward to it)
Although I haven't tried 99% of your work yet. But there's more to come!
I've started testing the results on different models. The results are often very interesting!
Very nice Alcina model, thanks for it.
On question, you put "mature woman" as a suggested tag for her. Yet this isn't a danbooru tag for as far as the Danbooru one is "mature female". ILXL models have somewhat of a limited understand of natural language. So it can understand "mature woman" yet this remains a bet. While using "mature female", a danbooru tag that all ILXL have been trained on, is safer when asking for any ILXL clip (OpenCLIP ViT-bigG/14 or CLIP ViT-L/14) to "understand" what it means. So have you trained this LoRA on "mature woman"? Otheriwse why suggesting it instead of "mature female", a Danbooru tag?
there just suggestions not a requirment, if mature female, is the better option or makes a difference then ill just put mature female, instead. But no I didn't specially tag mature female or woman in the dataset. I merge these terms into the TW if possible. its just something that can be added that may enhance the image or not depending on whatever model folks are using
@True_Might Thanks for the clarification.