lora trained from 185 hand-picked and signature/text removed images of artist LAM (https://www.pixiv.net/en/users/17429)
weight at `0.6` for less style, enable to mix with other styles. `0.8` - `1` make the style closer to artist's.
recommended tags:
`colorful, multicolored hair, multicolored eyes, eyeshadow, makeup`
nice to play with:
`gradient eyes`
model used for training is `anything-v4.5`
model for previews is `AbyssOrangeMix2`
**parameters:**
masterpiece, best quality, 1girl, solo, highres, <lora:lam-000010-v3:0.6>, upper body, portrait, colorful, multicolored hair, multicolored eyes, eyeshadow, makeup
Negative prompt: (worst quality, low quality:1.4), medium quality
Steps: 22, Sampler: DPM++ SDE, CFG scale: 6, Seed: 1394024…
