<h2>LoRA for Yuki-Onna the Ice Mayakashi, Yuki-Onna the Icicle Mayakashi, and Yuki-Onna the Absolute Zero Mayakashi</h2>I trained it with 128 pics, which was mainly fan art and some official art. I trained it at 20 epochs with a dim/alpha of 32/16 and with 2 steps.Please play around with the LoRA weights. Let me know how it goes for you. Read stuff below on guidelines and known issues :)<h2>Triggers:</h2><ul><li>Form 1 (ie: Ice Mayakashi) : input "YukiOnnaYGO, IceMayakashi, 1girl, pale skin" after the LoRA near the start of the prompt. Consider adding "bare thighs" or "bare legs".</li></ul><ul><li>Form 2 (ie: Icicle Mayakashi) : input "YukiOnnaYGO, IcicleMayakashi, 1girl, pale skin, white bow, sandals, medium breasts" after the LoRA near the start of the prompt. Consider adding "large breasts" or "black pantyhose".</li></ul><ul><li>Form 3 (ie: Absolute Zero Mayakashi) : input "YukiOnnaYGO, AbsoluteZeroMayakashi, 1girl, pale skin, sandals, medium breasts" after the LoRA near the start of the prompt. Consider adding "large breasts", "black pantyhose", "holding staff" and "hair ornament".</li></ul>NB: can place in negative prompts if the clothes are stubbornly appearing.<h2>Notes:</h2><ul><li>My experience has shown scale 0.8 ~ 1.1 works best; my best results were 1.</li></ul><ul><li>My prompt order shown in the sample pictures worked best for me.</li></ul><ul><li>Works on different models though weights may need to change for good results.</li></ul><ul><li>I used Anything4.5 model, Ambientmix, anyLora, and Midnight Maple model.</li></ul><ul><li>Refer to my pictures for examples and what upscaler I used, keep in mind I have used other textual inversions to help with poses and negatives.</li></ul>

Files with SHA256: 2aa0df051744b99ff4dfebfcbdf2fe8525342ebbd49046460dfca217a2d6d7b3