Violet Evergarden (Kyoto Animation / Kyoani) style LoRA
Making models can be expensive. Do you like what I do? Consider supporting me on Patreon 🅿️ or feel free to buy me a coffee ☕
There are a lot of nice LoRA nets about this amazing anime, but I still wanted to try making one. I took data from all the movies and the first season.
Should work fine at around weight 1 on AnyLoRA ckpt and AAM. Use violet evergarden to increase fidelity. If somehow you don't get flat colors on new concepts or lora combinations, use AAM.
Remember this is a style LoRA. To make characters accurately you need other networks too (but it can get pretty close with accurate descriptions).
How to use LoRA's in auto1111:
Update webui (use
git pulllike here or redownload it)Copy the file to
stable-diffusion-webui/models/loraSelect your LoRA like in this video
Make sure to change the weight (by default it's
:1which is usually too high)
Description
FAQ
Comments (10)
my god!! thank you very much, im waiting for this Lora for a long time
A little advice, Could you use the name of anime charactor to make the trigger words~?
no, this is not a character lora, it's a style lora. Its purpose is not to easily reproduce characters from the show, but to get the show artstyle and use it freely. If I tag characters, the model would pick characteristics off of character tags instead of descriptions, thus making the style less flexible and harder to apply to OCs.
If you want to make characters in this style, then combine it with a good, non overfitted, character lora/TI
Or use a style overfitted character lora. Your choice :)
@Lykon OK~I C~
Hello, sir. A little request for a style lora. Anime called Samurai Champloo :)
Wdym change weight around 1? You mean the offset number? Like 0.5 etc ? I'm kinda new to this
lora has a weight when you use it. See the video. Default is 1. "around 1" means 0.9 or 1.2 for example.
Hi, if possible can you import this model to tensor.art? I really want to try it there