Minori Yamato from the webcomic by Projekt Wolfenstahl "Deathblight". Works best around 0.7-1.0 weight. Training done on Pony Diffusion V6 XL. Request form

Trigger: "mnori, wolf girl, wolf tail, animal ears, multicolored eyes, black hair, large breasts, nail polish, toned" always needed for accuracy.

• Add "Hair ornament" for her classic hair ribbon that Ferania gave her in the comic.

• Add "abs, navel" where visible for proper physique.
  

• Hairstyles, choose one:

  • "short hair" for her main hairstyle.

  • "long hair" for her long hair from early chapters.

  • "single hair bun" for her hairstyle from earlier chapter.
    

Suggestions/notes:

• Halterneck dress Outfit: "purple halterneck, cleavage, purple dress, short dress, black gloves, fingerless gloves, elbow gloves, bike shorts, black thighhighs, black belt"

  • Somewhat inconsistent due to poor data but can work with some RNG.
    

• Plunging neckline Outfit: "purple dougi, plunging neckline, black choker, black gloves, fingerless gloves, elbow gloves, black thighhighs, belt bow"

  • The bow is a bit random but the data was poor sadly.

• Crop top Outfit: "multicolored crop top, japanese clothes, collarbone, sleeveless, midriff, black shorts"

  • Best used with "long hair" style as that was the original in the comic for the outfit.

  • Works well even for the limited data but the top has some RNG.
    

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

  • I suggest lower weight to 0.8 when using other outfits or loras.

  • The nails are slightly inconsistent but work better than when I made Ferania.

    • Remove "nail polish" if you dislike them or have issues with the nails.

Minori is best girl so it was a battle to get her working well, I managed to capture face, eyes, hair very well but sadly he outfits lacked good data so do not work as well as I would like. There is some style burn but sadly most of the data is from the comic and there was not much fanart to work with. I hope you all enjoy.

Feedback and reviews are always appreciated.

Training data is in the Lora metadata.

Trained on 105 images.