Sign In

Hyena Mayeso PonyXL (Deathblight | Crescentia Fortuna) | LoCon

47
173
573
22
Verified:
SafeTensor
Type
LyCORIS
Stats
173
573
151
Reviews
Published
May 29, 2024
Base Model
Pony
Training
Steps: 960
Epochs: 10
Usage Tips
Clip Skip: 2
Strength: 1
Trigger Words
hyemayeso, dark-skinned female, wolf girl, animal ears, freckles, wolf tail, multicolored tail, makeup, gold eyeliner, multicolored eyes, multicolored hair, streaked hair, medium hair, large breasts, toned
Low ponytail
pink nails, white nails, gradient nails
Abs, navel
Hash
AutoV2
4E1E1C549F
default creator card background decoration
Chameleon Badge
Manityro's Avatar
Manityro

Hyena Mayeso from the webcomic by Projekt Wolfenstahl "Deathblight". Works best around 0.8-1.0 weight. Training done on Pony Diffusion V6 XL. Request form

Trigger: "hyemayeso, dark-skinned female, wolf girl, animal ears, freckles, wolf tail, multicolored tail, makeup, gold eyeliner, multicolored eyes, multicolored hair, streaked hair, medium hair, large breasts, toned" always needed.

  • Add "Low ponytail" for her main hairstyle or remove for loose casual style.

  • Add "pink nails, white nails, gradient nails" for her proper nails but beware nails are odd on PonyXL.

  • Add "Abs, navel" for proper definition where visible.

Suggestions/notes:

  • Check above samples for outfit ideas, her main outfits did not work very well as the data was very limited and not good for teaching em properly.

  • Very flexible and NSFW able as there was a lot of nude data and the outfit was very normal.

    • The nails are very inconsistent even after I lot of testing to get them working so I suggest trying a bit but not torturing yourself to get them working right.

A fit Choco woof with abs and freckles like man this some good stuff right here. I really love the designs from the Deathblight comic so I wanted to make a few of the characters but dang these are tricky and very tag heavy just due to their complex design but came out really nice. I hope you all enjoy.

Feedback and reviews are always appreciated.

Training data is in the Lora metadata.

Trained on 67 images.