BanniArt 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。
在 Civitai 上基于毛茸茸艺术家 BanniArt 训练的 LoRA
又称 Banni Art / ArtBanni
请通过以下链接支持这位艺术家:
http://www.furaffinity.net/user/banniart/
https://www.deviantart.com/banniart
始终使用标签 by banniart
此 LoRA 基于 Pony V6 版本 / Illustrious
尽管使用了 Pony V6 基础模型 / Illustrious,但在某些其他模型中使用时,图像可能会出现不一致,尤其是在 眼睛和手部。若在 Civitai 上使用,建议搭配一些 Embedding / Face Fix 以获得更好的效果。若使用 ComfyUI,最佳效果需搭配 Hirefix / DDetailer 来改善角色面部图像。
别忘了使用与 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。
为在 Civitai 上更好地使用 LoRA,我建议使用 VAE (SDXL VAE) 以及 Embedding (Pony PDXL Negative Embeddings) 以减少手部和眼睛细节的错误。
请先检查与您所选图像兼容的模型,某些模型会影响 LoRA 在最终图像生成中的效果,尤其对于如:毛茸茸、人类、写实图像等类型。因此,同一模型并不总适用于不同类型的图像。
若希望强制图像生成器使用您选择的 LoRA 权重,请在正向提示中使用以下标签:
Illustrious (v1.0)
<lora:BanniArt_Style_Lora_Illustrious:0.8>
Pony (v1.0)
<lora:BanniArt_Style_Lora:0.8>
注意:LoRA 名称末尾的数字(0.8)代表 LoRA 权重。在图像生成器中我们有滑块调节权重,但若您想“强制”使用,请在正向提示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: <lora:BanniArt_Style_Lora_Illustrious:0.8>
Pony (v1.0)
Civitai 滑块 = 0.8
正向提示: <lora:BanniArt_Style_Lora:0.8>
- 我添加了与艺术家最知名角色相关的标签,若希望生成接近该角色的图像,请务必输入角色名称。以下为每个角色的正向提示示例:
Tida
by banniart, furry, furry male, anthro, tida, equine, horse, animal ears, animal nose, snout, muscular male, brown body, brown fur, body fur, dark-skinned male, dark skin, bedroom eyes, narrowed eyes, short hair, black hair, green eyes, armband, accessory, mohawk, 1boy, male focus, solo, ass, bara, tail, nude, muscular male, looking at viewer, lying, smile, on stomach, on bed, bedroom, light particles, shadow, from side, from below
Diffuse Moose
by banniart, furry, furry male, anthro, diffuse Moose, deer, moose, new world deer, muscular male, muscular anthro, brown fur, brown body, mustache, hooves, horn, scut tail, short tail, muscular, muscular male, solo, male focus, 1boy, bara, bulge, nipples, furry, abs, pectorals, underwear, male underwear, thighs, navel, large pectorals, chest hair, furry male, short hair, thick thighs, bare pectorals, stomach, open clothes, open shirt, shirt, facial hair, navel hair, feet out of frame, animal ears, mature male, sitting, couch, night, window, room
为获得版本 1.0(Illustrious)的最佳效果,由于艺术家独特的风格以及图像训练时 AI 难以准确识别人脸,我推荐使用 ADetailer(TensorArt)/ Face Fix(Civitai)。若您使用本地图像生成器(ComfyUI / A1111 / Forge),我也推荐以下设置:
Face Fix(Civitai)
- 工作流:图像到图像高分辨率修复
- 去噪值:0.2 ~ 0.4
- VAE:SDXL VAE
超分辨率(ComfyUI / A1111 / Forge)
- 超分辨率模型:4x_foolhardy_Remacri / R-ESRGAN 4x+
- 高分辨率步数:20
- 去噪强度:0.5
ADetailer(TensorArt)
- 超分辨率 = 超分辨率模型:R-ESRGAN 4x+ / 高分辨率步数:20 / 去噪强度:0.5
- ADetailer = 检测模型置信度阈值:0.5
- 修复遮罩模糊:8
- 去噪强度:0.4
- 仅遮罩修复:勾选(true)
以下为初始配置,其他配置方式由您自由选择:
ILLUSTRIOUS (v1.0)
模型
Artofura | 毛茸茸混合 - Illustrious++ - v2.0
LoRA
BanniArt Style Lora - v1.0 (illustrious) - 0.5 ~ 1.0 (建议 0.8)
高级设置
CFG Scale - 3.0 ~ 9.0 (建议 5)
采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (建议 Euler a)
步数 - 20 ~ 60 (建议 30)
Clip Skip - 1/2
VAE - SDXL VAE
正向提示
masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, depth of field, scenery, volumetric lighting, by banniart, furry, furry male, anthro, tiger, feline, animal ears, animal nose, snout, grey eyes, tiger stripes, grey fur, body fur, grey hair, short hair, male focus, 1boy, pectorals, solo, pants, braid, shirt, belt, partially unbuttoned, pectoral cleavage, looking at viewer, couch, white pants, sitting, long hair, collared shirt, fingernails, tiger tail, facial hair, single braid, red shirt, object in mouth, smoke, cigarette, cigarette in mouth, leaning back, black belt, mature male, indoors, collarbone, braided ponytail, large pectorals, feet out of frame, on couch, bangs, parted lips, bulge
反向提示
bad quality, worst quality, worst detail, sketch, censor
PONY (v1.0)
模型
Pony Diffusion V6 XL - V6 (从这个开始)
LoRA
BanniArt Style Lora - v1.0 (Pony) - 0.5 ~ 1.0 (建议 0.8)
高级设置
CFG Scale - 3.0 ~ 9.0 (建议 3)
采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (建议 DDIM)
步数 - 20 ~ 60 (建议 60)
Clip Skip - 1/2
VAE - 无
正向提示
score_9, score_8_up, score_7_up, score_6_up, by banniart, golden retriever, retriever, domestic dog, canine, canid, canis, solo, ass, male focus, 1boy, furry, nude, muscular, bara, completely nude, muscular male, ass focus, furry male, blue eyes, thick thighs, tail, from behind, thighs, hands on own ass, looking at viewer, animal ears, back, grabbing own ass, looking back, ass support, ass grab, smile, huge ass, gradient background, blue background, white background, low-angle view, dutch angle
反向提示
score_5_up, score_4_up, zPDXL3, 3D, 2D, human
由于该模型基于 Pony Diffusion,反向提示中无需过多指定,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:
在使用此 LoRA 时,若可能,请注明作者 BanniArt 的署名。
请勿出售或商业化此 LoRA。
请勿删除由 LoRA 生成图像的元数据。










