Shirokoma Style Lora (Illustrious / Pony)

详情

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,基于兽迷艺术家Shirokoma在Civitai上训练

也被称为白狛 / sirokoma / shirokoma_bgl / bungalow / shirokoma_bgls

请通过以下链接支持这位艺术家:

https://www.patreon.com/user?u=5318790

https://twitter.com/shirokoma_bgl

https://x.com/shirokoma_bgls

https://bungalow.booth.pm/

https://www.pixiv.net/en/users/2305461

  1. 始终使用标签 by shirokoma

  2. 此LoRA基于Pony V6版本 / Illustrious

  3. 尽管使用了Pony V6基础模型,但在其他模型中使用时,图像可能存在一些不一致,尤其是在眼睛和手部表现上。若在Civitai上使用,建议搭配一些Embedding / Face Fix以获得更好效果。若使用ComfyUI,最佳效果需搭配Hirefix / DDetailer来优化角色面部细节。

  4. 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,
  1. 在Civitai上调整LoRA权重由您自行决定。

  2. 若希望生成彩色图像,请将monochrome, greyscale, sketch标签加入负面提示中,因为该LoRA是基于黑白图像或单一色系训练而成。

  3. 若希望生成无任何屏蔽的图像,请将censored, mosaic censoring, bar censor, heart censor, convenient censoring标签加入负面提示中,因为部分由该LoRA生成的图像原作中包含屏蔽内容。

  4. 若希望强制图像生成器使用您设定的LoRA权重,请在正向提示中加入以下标签:

Illustrious (v1.0)

<lora:Shirokoma_Style_Lora_Illustrious:1>

Pony (v1.0)

<lora:Shirokoma_Style_Lora:1>

注意: LoRA名称末尾的数字(1)代表LoRA权重。在图像生成器中我们有滑块调节权重,但若想“强制”使用特定权重,请在正向提示中使用与生成器滑块相同的数值,例如:

Illustrious (v1.0)

Civitai滑块 = 1

正向提示: <lora:Shirokoma_Style_Lora_Illustrious:1>

Pony (v1.0)

Civitai滑块 = 1

正向提示: <lora:Shirokoma_Style_Lora:1>

注意: 该艺术家使用了多个知名兽迷角色(卡通、电影、漫画等),但在训练时我保留了LoRA的灵活性,使其可与其他LoRA混合使用。我的建议是使用特定角色的LoRA,以获得更满意的最终效果。

以下为初始配置,但您也可自由尝试其他配置:

ILLUSTRIOUS (v1.0)

模型

Nova Furry XL - Illustrious v6B

LoRA

Shirokoma Style Lora - v1.0 (illustrious) - 0.5 ~ 1.0 (建议使用1)

高级设置

CFG Scale - 3.0 ~ 9.0 (建议使用7)

采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (建议使用Euler a)

步数 - 20 ~ 60 (建议使用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 shirokoma, furry, furry female, anthro, lion, felid, 1girl, solo, beach, ocean, water, sky, day, clouds, sun, areola slip, navel, red slingshot swimsuit, tail, animal ears, arms behind head, body fur, yellow eyes, covered erect nipples, huge breasts, snout, pubic hair, sparkle, arms up, simple background, red bikini, from below, open mouth, teeth, female pubic hair, looking at viewer, tan fur, tan body, smile, micro bikini, standing, armpits, large breasts, large areolae, red one-piece swimsuit, animal nose, stomach, :3, thick thighs, nipple slip, sidelocks, long hair, blonde hair, cowboy shot, lion tail, pubic hair peek, wide hips, blush, glistening body, shiny body

负面提示

bad quality, worst quality, worst detail, bad quality, worst quality, worst detail, sketch, censor, human, multiple tails, modern, recent, old, oldest, graphic, cartoon, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digit, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, username, simple background, conjoined, ai-generated, worst quality, low quality, off-topic, old, early, chromatic aberration, abstract, jpeg artifacts, scan artifacts, logo, censor, censored, mosaic censoring, bar censor, heart censor, convenient censoring, monochrome, greyscale, sketch

PONY (v1.0)

模型

Pony Diffusion V6 XL - V6(请从这个开始)

LoRA

Shirokoma Style Lora - v1.0 (Pony) - 0.5 ~ 1.0 (建议使用1)

高级设置

CFG Scale - 3.0 ~ 9.0 (建议使用5)

采样器 - DDIM / Euler a / Euler / Heun / DPM2 / DPM++ 2M Karras / Uni PC (建议使用Euler a)

步数 - 20 ~ 60 (建议使用30)

Clip Skip - 1/2

VAE - SDXL VAE

正向提示

score_9, score_8_up, score_7_up, by shirokoma, domestic dog, canine, canid, 1girl, swimsuit, breasts, furry female, bikini, furry, animal ears, gold bikini, solo, tail, multicolored hair, yellow eyes, huge breasts, arms behind head, dark skin, armpits, blonde hair, body fur, brown fur, dark-skinned female, looking at viewer, side-tie bikini bottom, long hair, brown hair, arms up, cleavage, animal nose, bangs, smile, yellow bikini, gradient background, gray background, whute background 

负面提示

zPDXL3, 3D, 2D, human, monochrome, greyscale, sketch, censored, mosaic censoring, bar censor, heart censor, convenient censoring

由于此模型基于Pony Diffusion,负面提示中无需过多设定,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:

  • 使用此LoRA时,如有可能,请注明作者Shirokoma的署名。

  • 不得销售或商业化此LoRA。

  • 不得删除由该LoRA生成图像的元数据。

此模型生成的图像

未找到图像。