"And you don't seem to understand...."A lora based on Akimasa Sho's works,trained on NVIDIA V800 with 186 public pictures on author‘s pixiv page.Trained under 1024x1024 resolotion.DO NOT USE hi-res as it will cause damaged result. Use "lofi" promot to activite this lora."やっぱり……いい女には、たばこと屋上と……そしてヒラヒラ服だわね……"<h3 id="important-notice-i18ncfhy4">---IMPORTANT NOTICE---</h3>I noticed that this version is not as stylized as the previous one. So if you are looking for a stable way to produce high-resolution images, please choose v1.2, and if you want highly stylized stunning images, please choose v1.1 (for use with less than 1024 pixels width/height).<h3 id="copyright-warning-9c66rm40q">---COPYRIGHT WARNING---</h3>The copyright of these images belongs to the original author. The model was made for my own interest. Please do not use this lora for commercial purposes. If this model infringes your rights, please contact me directly in a private message and I will delete it within 72 hours.

Lofi-Akimasa Sho(あきま)

With improved source data and using a higher resolution (640*640) for training, this version should perform better on images with slightly larger resolutions.And this version of Lora is extremely stylized.

Files with SHA256: 3e6a1fb2fd8e6e35f6644d5759bb1d874c501185a39bf9ecd25da7dde2fb5981