Weight: around 0.7-1.0 for v2face and v1; 0.4-0.7 for v2dress
Works fine with both anime and realistic (tried on deliberate2)
For other LoRAs better to use v2face version than v2dress. Theres some training image examples on the v2dress tab.
Main costume: pink hair, buns, blue jacket, green skirt
School uniform: pink hair, buns, white shirt, sweater vest, green skirt
Happy sugar life - The series is centered on high school girl Satō Matsuzaka (Satou Matsuzaka) who befriends a mysterious little girl named Shio, who was abandoned by her mother, and immediately becomes highly attached to her. They both agree to live together in Satō's apartment, and the former vows to protect that feeling of love, even if it means committing crimes or even killing people.
Description
Anythingv3, 768x768, 4/16 epoch
i like faces on this epoch
FAQ
Comments (5)
Wanted to share some helpful tips I have found for using this lora, and others like it! This is a very beginner friendly guide. This works best for loras that actually have trigger words for their face and hair.
1. Instead of combining the lora full strength (using the actual lora file and weight in the initial prompt), start by describing Satou by using her trigger words: 1girl, Satou, long pink hair, hair buns, red eyes. when you first generate her, she will have too much detail and shading and her eyes won't be "hers", but that's perfectly fine. For the negatives, use the textual inversion embedding "easynegative" (it's searchable on civitai), and add other information you may need if Stable Diffusion is having any trouble. Initially, it may do weird stuff to the hair like make it short and give her twintails, so I suggest these in the negatives.
2. Next, add any other factors that are important, such as another lora to change her outfit. Make sure to add in all the environmental information you need as well.
3. Send it to inpaint and paint just her face. Use the lora file name at max strength in the positive prompts, and also add: Satou, red eyes, (expression you want her to make), high quality, high resolution, masterpiece. In the negatives, add things related to correcting blurriness and resolution issues.
4. Send your new image with the corrected face to image to image with your entire prompt copied over, even the lora file names for Satou and the outfit lora, both at full strength. When you work with this, simply change the denoising strength down to as low as .2, or as high as .35. If she still has too much shading and detail, add these to your negative prompts: detailed shading, soft shading, realistic, realism. For positives: anime shading, cel shading, low detail, anime style. with all of this plus the perfect base image, you can now have the Satou wear any outfit you need her to, and it'll be high quality and look the way the beautiful example images for those loras look! In the prep stages before adding the file name with full strength, it also listens to the colors much better, so if you struggled to get her to wear a certain outfit in a certain color (or it kept changing it to pink like her hair or something), this will correct that.
If the lora you want to use doesn't describe the face and hair, try using the lora on a lower strength and getting something closer to how the character looks, attempt to inpaint the face once you get something right, and then follow the same instructions using image to image on .2-.35 with max strength and it should fix it.
Details
Available On (2 platforms)
Same model published on other platforms. May have additional downloads or version variants.








