<h2>Description</h2>Since there is less fanart of Yurisaki Mika, I trained this LoRA based on Anything V4.5. Also because of it, overfitting is seriously occured while using this model. Including but not limit to the following would happend:<ul><li>Gesture will be the same if you have low weight of gesture prompt or it even doesn't exist (You may use ControlNet to fix that)</li><li>Pose may be strange</li></ul><h2>Limitations</h2>And limited to the limited train dataset, the following characteristics may be not accurate:<ul><li>Halo</li><li>Sailor collar</li><li>Number of flower ornaments</li><li>Number of wings</li></ul><h2>Hints</h2>Trigger word is `yurisaki_mika`, and the recommended weight is 0.6~0.8. Here are some prompts that help you generate more accuratly:<ul><li>`mika_serafuku`: specify the costume, Heaven 3 Mika</li><li>`mika_halo`: sometimes will generate a good Mika's halo</li><li>`low_twintails`: seldom is needed</li><li>`multiple_wings`</li><li>`beret`</li></ul>Surely you can add some negative prompt to avoid prompt polluting, or use some other prompts to specify Mika's costume. Generlization has been tested.A sample prompt using Counterfeit V2.5 with VAE:((masterpiece,best quality)), 1girl, (close_up), head_sculpture, solo, (low twintails:1.1), light_smile, &lt;lora:yurisaki_mika_20:0.8&gt;, yurisaki_mika, mika_serafuku, mika_halo, beret, multiple_wings, blue_sky, cloudsNegative prompt: EasyNegative, extra fingers,fewer fingersSteps: 28, Sampler: DPM++ 2M Karras, CFG scale: 6, Seed: 167769193, Size: 512x512, Model hash: a074b8864e, Denoising strength: 0.6, Clip skip: 2, ENSD: 31337, Hires upscale: 1.75, Hires steps: 20, Hires upscaler: R-ESRGAN 4x+ Anime6B<img src="https://imagecache.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/19e35f0a-ffa0-4ae7-244b-851a2b5b8200/width=525/19e35f0a-ffa0-4ae7-244b-851a2b5b8200" />This model was tested on Anything V4.5, Counterfeit V2.5, AOM V3, PastelMix, and all perform well as except.

Files with SHA256: ce49080006bdb370c38b76efeaefdd7ac7873ff5ed0f863a9b686d5500903ce4