Destabilizer 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,基于毛茸茸艺术家 Destabilizer 在 Civitai 上训练

或称为 destabilizer2

通过以下链接支持这位艺术家:

https://www.furaffinity.net/user/destabilizer

https://twitter.com/Destabilizer2

https://www.pixiv.net/users/37651878

https://inkbunny.net/Destabilizer

https://subscribestar.adult/destabilizer

  1. 始终使用标签 by destabilizer

  2. 此 LoRA 基于 Pony V6 版本 / Illustrious

  3. 尽管使用了 Pony V6 基础模型 / Illustrious,但在某些其他模型中,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,建议添加一些 Embedding / Face Fix 以获得更好效果。对于使用 ComfyUI 的用户,最佳效果是配合 Hirefix / DDetailer 来改善角色面部图像。

  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. 如果您希望生成彩色图像,请在负面提示中添加 monochrome, greyscale 标签,因为该 LoRA 是基于黑白图像或单一配色方案训练的。

  3. 首先检查与您选择的图像兼容的模型,某些模型会影响 Lora 在最终图像生成中的效果,例如:毛茸茸、人类、写实图像等。因此,同一模型并不总是适用于不同类型的图像。

  4. 如果您希望强制图像生成器使用您选择的 LoRA 权重,请在正向提示中添加以下标签:

Illustrious (v1.0)

<lora:Destabilizer_Style_Lora_Illustrious:1>

Pony (v1.0)

<lora:Destabilizer_Style_Lora-000009:1>

注意:LoRA 名称末尾的数字(1)表示 LoRA 权重。在图像生成器中我们有滑块调节权重,但如果您想“强制”使用,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:Destabilizer_Style_Lora_Illustrious:1>

Pony (v1.0)

Civitai 滑块 = 1

正向提示: <lora:Destabilizer_Style_Lora-000009:1>
  1. 我添加了与艺术家最知名角色相关的标签,如果您希望生成接近该角色的图像,请不要忘记输入角色名称,我将在每个角色下方提供正向提示示例:

Solii

by destabilizer, furry, furry female, anthro, solii, bovine, cattle, dark yellow horn, big areola, huge breasts, brown areola, brown body, brown fur, brown nipples, green eyes, narrowed eyes, grey hair, huge breasts, tail tuft, tail, thick thighs, tuft, 1girl, solo, lying, on back, on floor, wooden floor, looking at viewer, completely nude, above view, black pussy, seducing, seduction, legs together, arms above head

Molly

by destabilizer, furry, furry female, anthro, molly, equine, horse, animal ears, animal nose, snout, big areola, huge breasts, brown fur, white fur, body fur, two tone fur, brown pussy, long hair, blue eyes, white hair, curvy figure, voluptuous, 1girl, solo, sitting, couch, sofa, grabbing own breast, sweat, sweaty, sweat dripping, glistening body, shiny body, feet out of frame, looking at viewer

Shannon

by destabilizer, furry, furry female, anthro, shannon, rhinoceros, animal nose, animal ears, snout, huge ass, black hair, black nipples, (black sclera, yellow eyes:1.2), grey fur, body fur, hairy pussy, mohawk, nipple piercing, nose horn, nose ring, piercing, ring piercing, tail tuft, tattoo on arm, 1girl, solo, smile, looking back, looking at viewer, huge ass, presenting, presenting ass, raised tail, trees, savanna, leaning on tree, blurry background

注意:艺术家使用了多个广为人知的毛茸茸角色(卡通、电影、漫画等),但我在训练中使 LoRA 保持灵活,可与其他 LoRA 混合使用。我的建议是:使用特定角色的 LoRA,以获得更满意的结果。

以下是初始配置,但其他任何配置均可由您自行决定:

ILLUSTRIOUS (v1.0)

模型

WAI-NSFW-illustrious-SDXL - v14

LoRA

Destabilizer 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 destabilizer, furry, furry female, anthro, pokemon, generation 4 pokemon, lagomorph, leporid, lopunny, pokemon (species), animal nose, pink snout, animal ears, black sclera, red eyes, brown fur, body fur, 1girl, solo, huge breasts, solo, underboob, jeans, denim, shirt, pokemon \(creature\), rabbit ears, smile, thick thighs, black shirt, covered erect nipples, navel, looking at viewer, pregnant, crop top, standing, hand up, rabbit girl, big belly, open mouth, tongue, teeth, wide hips, midriff, dark blue pants, waving, stomach, blue background, gradient background, simple background

反向提示

bad quality, worst quality, worst detail, sketch, censor

PONY (v1.0)

模型

Pony Diffusion V6 XL - V6(建议从这个开始)

LoRA

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

高级设置

CFG Scale - 5.0 ~ 9.0 (建议使用 8)

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

步数 - 20 ~ 50 (建议使用 30)

Clip Skip - 1/2

VAE - SDXL VAE

正向提示

score_9, score_8_up, score_7_up, solo, 1girl, brown fur, ass, furry, black sclera, anus, from behind, colored sclera, furry female, pussy, rabbit tail, rabbit girl, simple background, tail, looking back, pokemon (creature), smile, body fur, ass focus, red eyes, two-tone fur, yellow fur, animal ears, rabbit ears, happy, pussy juice, thighs, white background, by destabilizer, lopunny, bunny, lagomorph, leporid, rabbit, pokémon

反向提示

score_6, score_5_up, score_4_up, zPDXL-neg, monochrome, greyscale

由于此模型基于 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:

  • 使用本 LoRA 时,如有可能,请注明作者 Destabilizer 的署名。

  • 不得出售或商业化此 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。