Loading
Du hast einen ComfyUI-Workflow? Erstelle direkt im Browser eine verständliche Workflow-Dokumentation.Workflow verständlich machen.Jetzt ausprobieren ↗
This is final ver ,about 2600 images of her from several sites,and after clean and filter, 1000 iamges expand to 2321 images for train.
Modellquelle
Quellenbeschreibung
Trigger Word :ai hoshino
This is final ver ,about 2600 images of her from several sites,and after clean and filter, 1000 iamges expand to 2321 images for train.
终稿,使用了几个图站上所有该角色的图共2600张,清洗筛选后1000张图扩增至2321张放入训练
please us v2 lora,v1 is worthless.
recommand use prompt:
aihoshino eyes,aihoshino hair ornament, purple hair,purple spank eyes
train base sdxl model.
recommend use kohakuXL_alpha7 model,without a doubt that is best anime sytle model I've ever used.
Quellen
1 QuelleModellartefakte
1 Artefakthoshinoai_v10_out.safetensors
safetensors · 218 MB · SHA-256 c0a19ad13de2…f936 · Civitai
HerunterladenQuellenauszüge
1 AuszugSource context: 1094 downloads · Type LoCon · Base model SDXL 1.0
recommend lora block weight 0,0,0,0,1,1,1,1,1,1,1,1 ,but it's not very useful,just better than nothing.
v2 lora lora block weight no test, can test it themselves
