The is a style lora, it does a lot of things and it was pretty difficult to make. Probably the hardest one I've made. It was created with a synthetic curated blend of images, meticulously hand picked.What my style Lora does, it changes the art style, improves the line work (in my opinion), add details, fixes/flattens colors, cleans images, adds better light and shadows.Few things to note:It was created mostly for NSFW stuff, you can still use it for SFW it does have enough training data but it's not what is specializes in.This isn't a checkpoint, compared to some of the highly stylized/improved ones, this probably falls short. But using this lora together with some of those checkpoints provides some interesting results, you will have to keep you weight lower than recommended in order to achieve this effect.Only MSRD is necessary for the style, but 1girl and blush were trained a lot into it all as well, it's all in the metadata for the lora if you want to look at it. Trained at 1024x1024. It does seem to favor woman (all that good nsfw stuff) and red eyes, if you don't want those add it to your negative prompt, no humans seems to work well. It was made to be flexible so it can be used with your masterpiece, good quality, and whatever other positive prompts you want.Recommended weight is somewhere around 0.7-0.9, you can use it at 1.0 but it can be a little unpredictable and struggle with the eye generations. 30+ steps seems to add the most details.All in all, I think it turned out pretty well :) let me know what you think or if you have any suggestions.

Files with SHA256: 5bd85a8103be11679052748a7ff2f08f37e92ba65a9299b95c477995c16cc19d