Hello everyone,
this is one of my first training attempts of a character-lora from Sayori Wakaba from the anime "Vampire Knight".
Prompt: look at "Trigger Words" on the right side 👉
Changelog: look at "About this Version" on the right side 👉
quality tags i use:
oekaki, anime coloring, anime screencap, masterpiece, best quality, amazing quality
Note: only ~30 trainings pics, so the model might be a bit unstable (e.g detail error with hands (like "crossed hands") and fingers)
My Settings (forge):
Checkpoint: waiNSFWIllustrious_v140
Sampler: Euler a
Scheduler: Automatic
Samplings Steps: 30
Hires.fix: R-ERSGAN 4x+Anime6B, 20 Hires steps, Denoising Strength: 0.5
Extentions (nice to have for little more Details):
Adetailer (highly recommended)
CFGRescale Integrated (forge)
CD Tuner
Detail Daemon
DynamicThresholding
FreeU
SelfAttentionGuidance
Description
upscaled training-images, delete subtitles, refined tag, change character Trigger to match danbooru tag (BUT (again) THIS model will not benefit from it, because the character hasn't a danbooru tag)