Dogyd Style Lora (Illustrious)

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 Dogyd(又名 syjo)在 Civitai 上训练

通过以下链接支持该艺术家:

https://e621.net/users/169136

  1. 始终使用标签 by dogyd

  2. 此 LoRA 基于 Illustrious 模型

  3. 尽管使用了 Illustrious 基础模型,但在某些其他模型中生成图像时,眼睛和手部可能出现不一致,尤其在 Civitai 上使用时,建议使用一些 Embedding / Face Fix 以获得更佳效果。对于使用 ComfyUI 的用户,最佳效果需搭配 Hirefix / DDetailer 来改善角色面部细节。

  4. 在 Civitai 上调整 LoRA 权重由您自行决定

  5. 为更好地在 Civitai 中使用 LoRA,我推荐使用 VAE (SDXL VAE) 以减少手部和眼睛细节的错误

  6. 若您希望生成彩色图像,请将 monochrome, greyscale, sketch 标签加入负面提示词,因为该 LoRA 是基于黑白图像或单一色系训练的

  7. 请先检查与您所选图像风格兼容的模型,部分模型会影响 LoRA 在最终图像生成中的效果,尤其针对:兽迷、人类、写实图像等类型。因此,同一模型未必适用于所有类型的图像

  8. 若您希望强制图像生成器使用您选择的 LoRA 权重,请在正向提示词中添加以下标签:

Illustrious (v1.0)

<lora:Dogyd_Style_Lora_Illustrious:1>

注意:LoRA 名称末尾的数字(1)代表 LoRA 权重。在图像生成器中,我们有权重滑块,但若想“强制”使用,请在正向提示词中输入与生成器滑块相同的数值,例如:

Illustrious (v1.0)

Civitai 滑块 = 1

正向提示词: <lora:Dogyd_Style_Lora_Illustrious:1>
  1. 我加入了与该艺术家最知名角色相关的标签,若您希望生成接近这些角色的图像,请不要忘记输入角色名称。以下为每个角色的正向提示词示例

Li Li Stormstout

by dogyd, li li stormstout, warcraft, bear, pandaren, furry, furry female, anthro, black hair, green eyes, black nose, black fur, black body, white fur, white body, body fur, ponytail, 1girl, solo, smile, staff, cowboy shot, long hair, chinese clothing, east asian clothing, pink kimono, ong sleeves, dark green sash, pink pants, baggy pants, female focus, temple, outdoors, blurry background, hand on hip, standing, puffy pants, closed mouth, smile, looking at side, holding staff, weapon,jacket, animal ears, pose action

Hilda (Warcraft)

by dogyd, hilda (warcraft), warcraft, canid, werecanid, werecreature, worgen, furry, furry female, anthro, brown hair, yellow eyes, black nose, brown fur, brown body, body fur, tuft, inner ear fluff, cheek tuft, claws, monochrome, greyscale, 1girl, solo, medium breasts, animal ears, female focus, long hair, suspenders, gradient background, closed mouth, looking at viewer, smile, fangs

以下为初始配置,但您可自由选择其他配置:

ILLUSTRIOUS

模型

NoobAI-XL (NAI-XL) - V-Pred-1.0-Version

LoRA

Dogyd 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 5.5)

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

步数 - 20 ~ 60 (建议使用 32)

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 dogyd, furry, furry female, anthro, crash bandicoot (series), coco bandicoot, bandicoot, marsupial, 1girl, solo, (mature female:1.4), flower, green eyes, hair flower, hair ornament, hand on hip, open mouth, teeth, bodysuit tight, light red bodysuit, long hair, cowboy shot, blue flower, covered navel, blonde hair, standing, medium breasts, big hair, pointy ears, smile, shiny bodysuit, contrapposto, two-tone fur, happy, body fur, turtleneck, short sleeves, hand up, animal ears, looking at viewer, legs apart, high ponytail, colored skin, covered collarbone, very long hair, ponytail, curly hair, animal nose, yellow fur, taut clothes, tail, cropped legs, wide hips

反向提示词

monochrome, greyscale, sketch, 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

通用建议:

  • 如可能,请在使用此 LoRA 时注明作者 Dogyd 的署名

  • 不得出售或商业利用此 LoRA

  • 不得移除由该 LoRA 生成图像的元数据

此模型生成的图像

未找到图像。