rh-z-image-base-lora
Description
| Field | Content |
|---|---|
| Model Name | rh-z-image-base-lora |
| Model Type | LoRA (text-to-image) |
| Platforms | ComfyUI / RunningHub / Hugging Face |
| Author | RunningHub-@AIGC-LONG |
| Original | https://www.runninghub.cn/model/public/2100968351605542914 |
| Try online | https://www.runninghub.ai |
| https://www.runninghub.cn/runninghub-api-doc-en/ |
About this model
# Z-Image Base｜ 极致光影｜增强补光｜超写实人像 LoRA
✅ 版本介绍
“写实人像光影增强”是Z-Image-Base微调LoRA模型V1.0版本的产物，旨在生成具有较强柔光风格滤镜的写实人像。该版本通过精细的参数调整和触发词的使用，成功地绘制出柔光镜风格高清写实人像,这个版本的训练集使用中英文双语打标，所以对中文和英文都有很好的理解能力。
✅ 核心效果：面部形成经典三角高光，低 - key 暗调画面，阴影层次细腻，强化人像立体感，适合艺术肖像、情绪写真创作，男女半身 / 胸像肖像 / 全身人像 通用。
✅ 核心特性
▸复调光影工程
伦勃朗三角光变体：45°左/右上入射
背景光域：暗调基底×高饱和彩光区
✅ 参数设置
▸ 触发词（Prompt）: rembrandtlight， 引导模型生成具有该模型独特的高清写实效果人像。
▸ 权重强度（Strength）: 0.7-0.95，用于控制图像的细节和风格化程度。
▸ CFG（Classifier-Free Guidance）: 6-9，用于增强图像的清晰度和风格一致性。
▸ 采样器（Sampler）: Euler，提供平滑的过渡效果。
▸ 调度器（Scheduler）: Simple，确保生成过程的高效性和稳定性。
▸ 推荐构图：1:1/2:3/3:4/9:16 , 半身 / 胸像肖像 / 全身人像，深色简约背景更容易出标准三角光效果。
⚠️ 提示：权重＞1 容易出现面部死黑、丢失皮肤细节，不建议拉高。
✅ 特点
▸ 色彩丰富: 图像中的色彩鲜艳且富有层次感，完美还原现实中色彩。
▸ 细节精致: 角色的服饰和装饰品细节丰富，展现出高超的艺术水准。
▸ 氛围独特: 背景中的场景与角色完美融合，营造出一种真实的光影氛围。
希望这个版本能够为你的创作带来更多的灵感和可能性！
正向示例提示词
rembrandtlight,best quality,ultra detailed,photorealistic,8k,raw photo, 1girl,solo,young east asian woman,upper body,looking at viewer,gentle expression,pure and innocent,lips slightly parted,natural makeup,pink lips, long brown hair,extremely long hair,straight hair,hair over shoulder,backlit hair,glowing hair strands,loose hair strands on face,wind-blown hair, flower crown,white flower wreath,white small flowers,green leaves,daisy crown, light green dress,mint green dress,sleeveless dress,tulle dress,chiffon dress,spaghetti strap dress,flowing dress, holding a green apple,green apple near face,hand holding apple,long white nails,french manicure, delicate silver bracelet,small pendant bracelet, outdoors,forest,woods,green trees background,grass,lush greenery,dappled sunlight through leaves, natural sunlight,side backlight,rim light on hair,bright and airy,soft diffused light,sunlit,fresh and clean, mint green color palette,low saturation,bright tone,ethereal,fairy-like,forest nymph,pure,japanese portrait style, medium close-up,vertical composition,shallow depth of field,bokeh,shot on 50mm f1.4,soft focus,film grain,fresh filter,
This repository provides rh-z-image-base-lora weight files. You can load them on RunningHub.
Finetuned from: Z-Image
Trigger words: rembrandtlight
Files
| File | Size | Purpose |
|---|---|---|
20260916-36209431_000001350.safetensors |
81 MiB | LoRA weights |
License
Published by RunningHub on behalf of the author. Copyright remains with the author. Follow the original project or upstream license.
Training at RunningHub
Train models on RunningHub: https://www.runninghub.ai/page-model
Run Seedance 2.5 with RunningHub API: https://www.runninghub.ai/call-api/api-detail/2133100000000700025
- RunningHub：https://www.runninghub.ai
- RunningHub 中国站：https://www.runninghub.cn
- RunningHub API 文档：https://www.runninghub.cn/runninghub-api-doc-cn/
- RunningHub API Documentation：https://www.runninghub.cn/runninghub-api-doc-en/
- Original project: https://www.runninghub.cn/model/public/2100968351605542914
- Author page: https://www.runninghub.cn/user-center/2034247767823753217