This is a super early build of this LORA! I only used 20 low resolution April Fool's images from the gallery to test if it would work, I'll update this model once I finish cropping and tagging the remaining 300 characters. Important -your checkpoint affect the color wash and sharpness (NAI works best for me) -It does not work well when mixing it with more than 2 Lora.-V0.2 ONLY improves 2:3 aspect ratio, it does not seem to be better than V0.1 as of my testings. -on Negative prompts you can use: jpeg artifacts to get a sharper image As of now it does not contain all Fate characters, so you need to either mix it with an existent fate character LORA or your checkpoint must already contain one. I tried it both with minimalist tags (like the Sonic example) and with a plentiful of tags using my OCs. I also tried mixing it with various LORA and even with no trigger words it overrides pretty drastically at a value of :1 there is still testing in progress but please give me your feedback in the comments on what you think I can improve on this model. I hope you enjoy it! :D (also do not use this lora for any commercial or copyrighted application, this is just a fan project I made because I love the artist's style, do not replicate it)

Fate April Fool's Riyo style (Lyomsnpmp)

-Increased the training resolution to 512x768, -added 20 more April Fools character to the training folder. -should work better for 2:3 aspect ratio canvas,-Improved hair styles, worsen eyes quality -JPEG ARTIFACTS NOT FIXED YET-

Files with SHA256: c7bca140eebecb6a293f58e1d423ed24477d6bd1cbaddbd39f9d2ec61ad3c1e2