薇塔 Vita_HI3 [XL]

低模ArtiWaifu Diffusion 以896x1280训练，支持任何常规分辨率（垂直或水平，只要为64的倍数即可）。推荐权重：<0.85> 至 <1>。

基本触发词 = Vita,long_hair,bangs,grey hair,purple eyes,yellow pupils

服装 = Vita_default_outfit,black_dress,pants,white_short_capelet,single long sleeve,single leg pantyhose,cleavage,breasts

细节 = glove,hair_ornament,jewelry,earrings

如何使用（已测试）：

1. 由于模型经过背景移除训练，因此背景部分可自由发挥。

2. 更换服装与细节时，无需添加提示词（若更换为与“dress”相关的服装，建议将权重调至<0.5>以下）。

3. 该模型可稳定配合大部分LoRA使用（适当降低权重）。

[Eng.ver]

Trained with ArtiWaifu Diffusion（896x1280） supports any standard resolution (vertical or horizontal, as long as it's a multiple of 64). Recommended weight: <0.85> to <1>.

Basic trigger words = Vita, long_hair, bangs, grey hair, purple eyes, yellow pupils

Outfit = Vita_default_outfit, black_dress, pants, white_short_capelet, single long sleeve, single leg pantyhose, cleavage, breasts

Details = glove, hair_ornament, jewelry, earrings

How to use (Tested):

1. Since it was trained with background removal, you're free to experiment with the background as desired.

2. For changing outfits and details, you can just omit the prompts (if switching to an outfit related to "dress," it's recommended to reduce the weight to <0.5>).

3. This model can be stably combined with most LoRA models (reduce weight accordingly).