
Trained on Anything V5, 13 images. This one is sometimes generated from a profile, you might have to use a pose prompt if you want a specific pose. Otherwise, I am extremely happy with the results.
Works great with PerfectDeliberate v10 and RuviStyle_v2.1 LoRas. I found these two in particular to produce both accurate results as well achieving that "cute" factor.
Other LoRas like ThorneMix, Dark Sushi 2.5D 大颗寿司2.5D, Coffee NSFW, darkSushi 2.5D_V2.0, Yotta Mix v5, and GrahamAingPikas-ClearVAE 2.3 for example should also produce interesting results.
I use Euler A
23 sampling steps ( I use this little because of testing, feel free to use more)
7 CFG
Thumbnail Model: PerfectDeliberate v10
https://pixai.art/model/1700714752332167596/1700718664563533469
描述:
PixAI made LoRa
训练词语: HinakoKominato, short hair, brown hair, brown eyes, white dress shirt, dark blue jacket, suit jacket, green skirt, average skirt, stockings, tiny breasts,
名称: 1700714752332167595.safetensors
大小 (KB): 38163
类型: Model
Pickle 扫描结果: Success
Pickle 扫描信息: No Pickle imports
病毒扫描结果: Success