Real Chinese Beauty

LORA
Original


Updated:

✨ Hey, guys~

This is a Realistic Enhanced Lora of Chinese female.

1️⃣ Characters with variable and unfixed face shapes.

2️⃣ Variety of scenes (daily life, vacation travel ......)

3️⃣ Rich composition (close-up, half-body, panorama can be).

🥳 I hope you like it and have a good time! If you can, please make more pictures...

❤️ [FLUX.1] versions

✅ Recommended Parameter

🖼️ Photo size:

1280*2048, 1024*1536, 1536*1024, 768*1024, 768*1152, 768*1280, 1024*1024, 1024*768, 1152*768, 1280*768,

you can use other image sized too!

⭕️ Sampler: uni_pc / Euler / dpmpp_2m

⭕️ Schedule: simple / beta

⭕️ Sampler steps: 20 ➡️ 60

⭕️ Guidance Scale: 3.5➡️10

⭕️ VAE: Automatic

----------------------------------------------------------------------------------

❤️ [ SD3.5 L ] versions

✅ recommended parameter

❌ Negative prompt:

bad hands,Extra fingers,Extra hands,blurry,noisy,unattractive,sloppy,ugly,(worst quality, low quality, normal quality:1.45),low-res,simple background,

🖼️ Photo Aspect Ratio:

768*1024, 768*1152, 768*1280, 1024*1024, 1024*768, 1152*768, 1280*768, 1024*1536

you can use other image sized too!

✅ ADetailer

⭕️ ADetailer Model: face_yolov8n_v2.pt

⭕️ Detection model confidence threshold: 0.25

⭕️ Denoising Strength: 0.35

❗️ ADetailer plug-in can be turned on as appropriate for close-up portraits, half and full body pictures are highly recommended!

----------------------------------------------------------------------------------

❤️ [ SDXL ] versions

✅ recommended parameter

❌Negative prompt:

(3d, render, cgi, doll, cartoon, painting, illustration, fake, ugly face, 3d modeling:1.4), bad hands,Extra fingers, Extra hands, blurry, noisy, unattractive, sloppy, ugly, (worst quality, low quality, normal quality:1.5), low-res, simple background, molten body, malformed body, bad anatomy, muscular, book, child, chibi, busty, mutated hands, low res, blurry face,pumped body, athletic body, black and white, text, watermark, artist name, signature,

🖼️ Photo Aspect Ratio:

768*1024, 768*1152, 768*1280, 1024*1024, 1024*768, 1152*768, 1280*768, 1024*1536

you can use other image sized too!

⭕️ Sampling Method:

DPM++ 3M SDE Karras

DPM++SDE Karras

DPM++ 3M SDE Exponential

Restart

DPM++ 2M Karras

⭕️ Sampler steps: 20 ➡️ 60

⭕️CFG: 3 ➡️ 10

⭕️VAE: none

✅ Hires.fix

⭕️ Hires upscaler: R-ESRGAN_4x ,4x_NMKD-Superscale-SP_178000_G,4x-UltraSharp

⭕️ Hires steps: 10 ➡️ 15

⭕️ Denoising: 0.3 ➡️ 0.5

✅ ADetailer

⭕️ ADetailer Model: face_yolov8n_v2.pt

⭕️ Detection model confidence threshold: 0.25

⭕️ Denoising Strength: 0.35

❗️ ADetailer plug-in can be turned on as appropriate for close-up portraits, half and full body pictures are highly recommended!

✨ 嘿,大家好~ 这是一款写实增强版的中国女性LoRA模型。

1️⃣ 人物脸型多变,没有固定模板。

2️⃣ 支持多种场景(日常生活、度假旅行……)

3️⃣ 构图丰富(可拍特写、半身、全景)

🥳 希望大家喜欢,玩得开心!如果可以的话,请多生成些图片呀……

❤️ [FLUX.1] 版本

✅ 推荐参数

🖼️ 图片尺寸:

1280×2048,1024×1536,1536×1024,768×1024,768×1152,768×1280,1024×1024,1024×768,1152×768,1280×768, 也可以使用其他图片尺寸!

⭕️ 采样器:uni_pc / Euler / dpmpp_2m

⭕️ 调度方式:simple / beta

⭕️ 采样步数:20 ➡️ 60

⭕️ 引导尺度:3.5➡️10

⭕️ VAE:自动

----------------------------------------------------------------------------------

❤️ [SDXL] 版本

✅ 推荐参数

❌ 负面提示词:

(3d, render, cgi, doll, cartoon, painting, illustration, fake, ugly face, 3d modeling:1.4), bad hands,Extra fingers, Extra hands, blurry, noisy, unattractive, sloppy, ugly, (worst quality, low quality, normal quality:1.5), low-res, simple background, molten body, malformed body, bad anatomy, muscular, book, child, chibi, busty, mutated hands, low res, blurry face,pumped body, athletic body, black and white, text, watermark, artist name, signature,

🖼️ 图片大小:

768×1024,768×1152,768×1280,1024×1024,1024×768,1152×768,1280×768,1024×1536 也可以使用其他图片尺寸!

⭕️ 采样方法: DPM++ 3M SDE Karras DPM++SDE Karras DPM++ 3M SDE Exponential Restart DPM++ 2M Karras

⭕️ 采样步数:20 ➡️ 60

⭕️ CFG值:3 ➡️ 10

⭕️ VAE:无

✅ 高清修复

⭕️ 高清放大算法:R-ESRGAN_4x、4x_NMKD-Superscale-SP_178000_G、4x-UltraSharp

⭕️ 高清修复步数:10 ➡️ 15

⭕️ 去噪强度:0.3 ➡️ 0.5

✅ ADetailer插件

⭕️ ADetailer模型:face_yolov8n_v2.pt

⭕️ 检测模型置信度阈值:0.25

⭕️ 去噪强度:0.35

❗️ 特写人像可适当开启ADetailer插件,强烈推荐半身和全身图使用!

Version Detail

SDXL 1.0

Project Permissions

Reprinting is strictly prohibited

    Use Permissions

  • Use in TENSOR Online

  • As a online training base model on TENSOR

  • Use without crediting me

  • Share merges of this model

  • Use different permissions on merges

    Commercial Use

  • Sell generated contents

  • Use on generation services

  • Sell this model or merges

Related Posts

Describe the image you want to generate, then press Enter to send.