1，本模型所有数据全来视频截图

2，最大出图大小建议不要超过 1280 x 1280

3，建议比重 在0.7-0.9左右脸和衣服还原不够提高比重，

4，采样方式 DPM＋＋2Mkarras   步数  20-30  重绘-0.3-0.5

5，算法 4×-UltraShrt   高分步数 15-20  

展示图描述词

(1girl:1.1),(looking at viewer:1.4),solo,(hair ornament:1.4),

(bangs:1.4),(purple hair:1.4),(long hair:1.6),(low twintails:1.4),(bell:1.4),(jingle bell:1.4),(hair bell:1.6),(animal ears:1.6),(fox ears:1.4),(animal ear fluff:1.4),(jewelry:1.6),(crystal:1.4),(facial mark:1.6),(forehead mark:1.6),(water_eyes:1.2),(decorations on the neck:1.4),(dress:1.6),(japanese clothes:1.6),(kimono:1.4),(miko:1.4),(red dress:1.4),(wide sleeves:1.4),(long sleeves:1.4),(skirt:1.4),(socks:1.4),(white socks:1.4),(shoes:1.1),(boots:1.1),(fluffy:1.3),(frilled:1.4),(lace:1.4),(necklace:1.4),(hair_ribbon:1.3),(ribbon:1.3),(spandex:1.4),(clothes fluttering in the air:1.4),(waterfall:1.5),(castle:1.1),(cityscape:1.6),(sky:1.4),(white clouds:1.6),(cherry_blossoms:1.3),(sakura tree:1.6),(river:1.4),(flowers and trees:1.3),(petals:1.5),(water:1.6),