Beat1213 Style Lora (Illustrious / Pony)
详情
下载文件
关于此版本
模型描述
请先阅读“关于此版本”部分中的版本详情和建议。
LoRA,由Civitai基于毛茸茸艺术家Beat1213训练
又称 12teab13 / びーと君 / beatkunganbaranai / beat-kun
请通过以下链接支持这位艺术家:
https://www.pixiv.net/en/users/6064906
始终使用标签 by beat1213
此LoRA基于Pony V6版本 / Illustrious
尽管使用了Pony V6基础模型,但在某些其他模型中生成图像时可能会出现不一致,尤其是在眼睛和手部。若在Civitai上使用,建议搭配一些Embedding / 面部修复以获得更好效果。若使用ComfyUI,最佳效果是搭配Hirefix / DDetailer来改善角色面部细节。
不要忘记使用与Pony模型相关的标签:
score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up,
在Civitai上调整LoRA权重由您自行决定。
为了在Civitai上更好地使用LoRA,我建议使用VAE (SDXL VAE) 和 Embedding (Pony PDXL 负向嵌入),以减少手部和眼睛的细节错误。
如果您希望生成彩色图像,请在负向提示词中加入 monochrome, greyscale, sketch 标签,因为此LoRA是基于黑白图像或单一配色方案训练的。
如果您希望生成无任何类型审查的图像,请在负向提示词中加入 censored, bar censor, out-of-frame censoring, mosaic censoring, convenient censoring, novelty censor, character censor 标签,因为部分由LoRA生成的图像原作者在绘制时带有某种审查。
请先检查与您选择的图像类型兼容的模型,某些模型会影响LoRA在最终图像生成中的效果,尤其对于毛茸茸、人类、写实图像等类型。因此,同一模型未必始终适用于不同类型的图像。
如果您想强制图像生成器使用您选择的LoRA权重,请在正向提示词中加入以下标签:
Illustrious (v1.0)
<lora:Beat1213_Style_Lora_Illustrious:0.8>
Pony (v1.0)
<lora:Beat1213_Style_Lora:0.8>
注意:LoRA名称末尾的数字(0.8)指的是LoRA权重。在图像生成器中,我们有权重滑块,但如果您想“强制”使用特定权重,请在正向提示词中使用与图像生成器滑块相同的数值,例如:
Illustrious (v1.0)
Civitai滑块 = 0.8
正向提示词: <lora:Beat1213_Style_Lora_Illustrious:0.8>
Pony (v1.0)
Civitai滑块 = 0.8
正向提示词: <lora:Beat1213_Style_Lora:0.8>
以下为初始配置,但其他任何配置均由您自行决定:
ILLUSTRIOUS
模型
Nova Furry XL - Illustrious v6B
LoRA
Beat1213 Style Lora - v1.0 (illustrious) - 0.5 ~ 1.0 (推荐 0.8)
高级设置
CFG Scale - 3.0 ~ 9.0 (推荐 7)
采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (推荐 Euler a)
步数 - 20 ~ 60 (推荐 30)
Clip Skip - 1/2
VAE - SDXL VAE
正向提示词
masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, depth of field, scenery, volumetric lighting, by beat1213, furry, furry male, anthro, equine, horse, nightclub, public, bulge, thick thighs, 1boy, male focus, muscular male, blush, smile, teeth, looking at viewer, yellow eyes, animal ears, fishnets, detached collar, cup, long hair, wine bottle, holding tray, wine glass, tail, drinking glass, bowtie, thighhighs, large pectorals, bow, animal ears, leotard, tail, breath, strapless, fishnet thighhighs, mature male, crossdressing, wrist cuffs, sweat, erection under clothes, strapless leotard
负向提示词
facial hair, beard, moustache, monochrome, greyscale, sketch, censored, bar censor, out-of-frame censoring, mosaic censoring, convenient censoring, novelty censor, character censor, bbad quality, worst quality, worst detail, sketch, censor
PONY (v1.0)
模型
Incursio's Meme Diffusion (SDXL, Pony) - v2.7PDXL
LoRA
Beat1213 Style Lora - 0.5 ~ 1.0 (推荐 0.8)
高级设置
CFG Scale - 3.0 ~ 9.0 (推荐 5)
采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (推荐 Euler a)
步数 - 20 ~ 60 (推荐 30)
Clip Skip - 1/2
VAE - SDXL VAE
正向提示词
score_9, score_8_up, score_7_up, by beat1213, bear, closed eyes, low-angle view, worm's-eye view, male focus, 1boy, ass focus, bara, furry male, solo, male underwear, bulge, bear ears, underwear, thighs, ass, furry, thick thighs, muscular male, underwear only, japanese clothes, brown fur, pectorals, white male underwear, animal ears, fundoshi, from below, muscular, white background, topless male, thick eyebrows, plump, dark skin
负向提示词
zPDXL3, 3D, 2D, monochrome, greyscale, sketch, censored, bar censor, out-of-frame censoring, mosaic censoring, convenient censoring, novelty censor, character censor, human
由于该模型基于Pony Diffusion,负向提示词中无需特别指定太多内容,因为Pony模型本身非常灵活,但我始终保留zPDXL-neg标签(/model/332646/pony-pdxl-negative-embeddings)。
通用建议:
如有可能,在使用此LoRA时请注明作者Beat1213的署名。
不得出售或商业使用此LoRA。
不得移除由LoRA生成图像的元数据。











