Prompt | illustration, (masterpiece:1.3), best quality, ultra-detailed, <lora:KudamakiTsukasaProject_kudamakiTsukasaV10:0.9>, upper body, solo, 1girl, kudamaki tsukasa, smile, closed mouth, looking at viewer, short-length hair, narrow hair between eyes, long and narrow bang, fox ears, yellow eyes, white overalls down to the upper arms and knees, blue sky, clouds, landscape scenery, trees |
---|---|
Negative prompt | EasyNegative, badhandv4, (worst quality, low quality:1.4), out of frame, realistic, lip, nose, tooth, rouge, lipstick, eyeshadow, abs, muscular, rib, (depth of field, bokeh, blurry:1.4), greyscale, text, watermark, logo, signature |
Sampler | DPM++ 2M Karras MOD |
Seed | 514301160 |
Steps | 30 |
Cfg Scale | 7 |
Model Hash | d124fc18f0 |
- ‹ prev Search: has:metadata next ›
I have never trained a LoRA before, but I recommend using Danbooru-style tags to keep the prompts as short as possible (especially for those that like to keep their LoRA files indefinitely or for non-Civitai extension users).
From Danbooru, Kudamaki Tsukasa is described to be a fox girl with brown eyes, short blonde hair, fox ears, and a tail. She wears a white romper with green ribbons, and white socks (which you already have down). Outside of that, she does have hair between her eyes, which you have included in your tags already.
The prompts should be something like...
kudamaki tsukasa, fox girl, short blonde hair, hair between eyes, brown eyes, white romper, green ribbons, white socks, fox tail
Sorry to go, "uhhh 🤓👆," but other than all that, this LoRA works very well as intended.