Hey everyone! 🌟 This is my very first try at making a style-based LoRA, and I’m so excited to share it with you! 🎨✨
By the way, I love the artwork by Mulerarts—her style is absolutely amazing! 💫 If you enjoy her work too, be sure to support her creativity! 🎀💖 Here’s a link to her Telegram channel: https://t.me/artsmuller. Let’s spread the love for awesome art! 🥰
Right now, my LoRA is still a work in progress—I’ll be improving it (goodbye watermarks, hello more training images! 🚀), but the results already look so cool that I just couldn’t wait to show you! 😍 Hope you like it! 💖
I’ve noticed that the style starts to show best with a LoRA weight of around 1.5–2.0. I’m not entirely sure why such a high value is needed, but next time, I’ll try to adjust the model so that it works well at a weight of 1.0. This LoRA performs best with anthropomorphic animals, though it handles humans decently too—just make sure to lower the weight slightly for human subjects.
Description
FAQ
Comments (4)
Finally style that isn't Eastern anime garbage, I can't even tell difference between them. now that is a distinct style. Too much furry but either way Good job man.
Thank you! There are so many furry images because the LoRA training used a lot of anthropomorphic character artwork. I'm still figuring out how to properly extract just the style from the images.
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.