This LoRa is aimed to somewhat recreate the style of the artist k1ko [https://linktr.ee/justk1ko] to varying degrees of success ; - ;
A good range is ( 0.65 to 0.70) :D
It's very good at making bottom heavy and feline like characters, anything else and you will have to play around with weighing prompts and negatives to get what you're after
VERSIONS:
V1: It doesn't match as closely to the the style but works great for most use cases since poses are less rigid
V1.5: It matches more closely to the style but somewhat more rigid
V2: It matches the style even more closely and is very forgiving!, only downside is that it's only workable on NovaFurryXL v5.0 and above, it can kind of work on Noob but there may be more glaring errors during generations
Stiff!
Depending on what you're doing i'd recommend adding the included negative tags because if you use the LoRa at a high strength (0.70<) the stiffness of the training data can be quite noticeable!. Mostly to take into consideration with V1.5
Will remove if asked