第五人格-啦啦队员/Identity V-cheerleader（small)

触发词： 1girl, backwards hat, orange hair, bangs, blue hair ribbon, low twintails, blue eyes, button eyes, white shirt, shirt tucked in, black shorts, black socks, black shoes, stitches, holding pom poms,

支持的风格：

官方图片风格：yuan style,2d style,

低面数游戏建模风格：game style, D5 style

旧装触发词：1girl, backwards hat, orange hair, bangs, blue hair ribbon, low twintails, blue eyes, button eyes, white shirt, shirt tucked in, black shorts, black socks, black shoes, stitches, holding pom poms,dirty clothes, torn clothes, dirty body, dirty face,

使用技巧：

1.在waiNSFWIllustrious_v140.safetensors上用32图训练。在它上面运行最好。

2.若生成全身图，在1080*1448的大分辨率且有ADetailer的小分辨率表现好，小分辨率且不开ADetailer可能导致脸部表现差。

3.这个lora的目的是生成更接近官方图风格的图片，如果你想要生成保证啦啦队员的基本特征的同时，不需要其它Lora就有丰富的风格。可以用我的另一个啦啦队员lora。

4.由于官方图作为训练图的限制，生成画面场景可能显得简单，在负面词添加“simple background,”可以缓解。或者可以用我的另一个啦啦队员lora。

English：

Trigger Words:

1girl, backwards hat, orange hair, bangs, blue hair ribbon, low twintails, blue eyes, button eyes, white shirt, shirt tucked in, black shorts, black socks, black shoes, stitches, holding pom poms,

Supported Styles:

1.Official art style: yuan style, 2d style

2.Low-poly game modeling style: game style, D5 style

Vintage Outfit Trigger Words:

1girl, backwards hat, orange hair, bangs, blue hair ribbon, low twintails, blue eyes, button eyes, white shirt, shirt tucked in, black shorts, black socks, black shoes, stitches, holding pom poms, dirty clothes, torn clothes, dirty body, dirty face,

Usage Tips:

1.Trained with 32 images on waiNSFWIllustrious_v140.safetensors. Performs best when run on this model.

2.For generating full-body images:

Excellent performance at a high resolution of 1080*1448 with ADetailer enabled.

Poor facial rendering may occur at low resolutions or when ADetailer is disabled.

3.The purpose of this LoRA is to generate images closer to the official art style. If you want to retain the cheerleader's core features while achieving diverse styles without other LoRAs, use my other cheerleader LoRA.

4.Due to limitations of using official images as training data, generated scenes may appear simplistic. This can be mitigated by adding "simple background," to the negative prompts. Alternatively, use my other cheerleader LoRA.