Loading
Du hast einen ComfyUI-Workflow? Erstelle direkt im Browser eine verständliche Workflow-Dokumentation.Workflow verständlich machen.Jetzt ausprobieren ↗
A LOCON for Stable Diffusion 1.5 of Rina Tennoji from Love Live. This model was trained and tested on the Mistoon Anime model using Kohya SS .
Modellquelle
Quellenbeschreibung
Overview
A LOCON for Stable Diffusion 1.5 of Rina Tennoji from Love Live. This model was trained and tested on the Mistoon Anime model using Kohya SS .
Note : My models are trained heavily to recreate not only the character but also their original art style. This makes the models more prone to overfitting, so they might not work well with other SD 1.5 models. You can check out the Epoches folder for alternative versions that might work better for your use-case.
Usage
Trigger keywords : eli ayase
Recommended keywords : blonde hair, blue eyes
Recommended strength : 0.6
Quellen
1 QuelleModellartefakte
1 ArtefaktRinaTennoji_Mistoon_Locon_CulturedDiffusion.safetensors
safetensors · 594 MB · SHA-256 16314f265b10…d025 · Civitai
HerunterladenQuellenauszüge
1 AuszugSource context: 444 downloads · Type LoCon · Base model SD 1.5
Recommended CFG : 4
Recommended SD model : Mistoon Anime model
Clip skip : 2
Rina-chan board
Trigger keywords : rina-chan board
Recommended keywords : sketchbook (if you want specifically the sketchbook version)
Rina-chan board Tips :
Increase the weight of rina-chan board (and sketchbook if you use it). They rarely work at weight 1.0.
Avoid tags that mention eyes (ex: yellow eyes and beautiful eyes ).
Avoid negative tags related to sketch / doodles .
Preferably avoid expression tags (ex: smile ). They can work, but from my tests the model is more likely to draw Rina's actual face instead of the board.
Some tags (like > ) seem to give a strong push toward drawing the board.
Experiment and share your results!
