license: other
datasets:
- p1atdev/pvc
language:
- en
library_name: diffusers
thumbnail: "https://s3.amazonaws.com/moonup/production/uploads/1677743056321-6305db1fcfbde33ef7d480ff.png"
tags:
- 文生图
- 稳定扩散
- 安全张量
widget:
- text: pvc, anime, masterpiece, best quality, exceptional, 1girl, bangs, bare shoulders, beret, black hair, black shorts, blue hair, bracelet, breasts, buttons, colored inner hair, double-breasted, eyewear removed, green headwear, green jacket, grey eyes, grey sky, hat, jacket, jewelry, long hair, looking at viewer, multicolored hair, neck ring, o-ring, off shoulder, rain, round eyewear, shorts, sidelocks, small breasts, solo, sunglasses, wavy hair, wet, zipper
example_title: WD1.5风格女孩
- text: pvc, anime, masterpiece, best quality, exceptional, 1girl, blonde hair, hat, baseball cap, aqua eyes, earrings, hoop earrings, yellow shirt, looking at viewer, upper body, simple background
example_title: 金发女孩
- text: pvc, masterpiece, best quality, exceptional, 1girl, cat ears, red hair, long hair, hairpin, swept bangs, yellow eyes, black jacket, white shirt, blue tie, white gloves, hand up, upper body, looking at viewer, buildings
example_title: 红发女孩
- text: nendoroid, masterpiece, best quality, exceptional, 1girl, cat ears, red hair, long hair, hairpin, swept bangs, yellow eyes, black jacket, white shirt, blue tie, white gloves, hand up, upper body, looking at viewer,
example_title: 粘土人风格
- text: figma, masterpiece, best quality, exceptional, 1girl, cat ears, red hair, long hair, hairpin, swept bangs, yellow eyes, black jacket, white shirt, blue tie, white gloves, hand up, upper body, looking at viewer, buildings
example_title: figma可动模型风格
PVC v3模型
本模型是基于Waifu Diffusion v1.5 beta 2微调的潜在扩散模型,专用于生成PVC手办风格的图像。您可以使用Danbooru标签体系来生成图像。
下载选项
建议使用WD的VAE模型以获得最佳效果!
同时可在负面提示词中使用badquality嵌入!
提示词指南
触发词
技巧
PVC手办风格更接近动漫风格而非写实风格。建议在正向提示词中添加anime
或在负面提示词中添加realistic
来优化效果。若想避免过于真实的面部表现,这个方法很有效!
示例图库
(此处保留原有示例图片及参数详情,仅翻译说明文字)
训练信息
参数 | 值 |
服务平台 | Runpod |
显卡型号 | A5000 |
训练脚本 | Linaqruf/kohya-trainer |
训练成本 | 约2美元 |
训练时长 | 约6小时 |
训练数据集 | 来自p1atdev/pvc的7467张图片 |
分辨率 | 896 |
训练轮次 | 5 |
优化器 | Lion |
学习率 | 4e-7 |
学习率调度 | 余弦退火重启 |
训练批大小 | 1 |
🤗 Diffusers使用指南
通过Diffusers库高效运行Stable Diffusion 2:
pip install diffusers transformers accelerate scipy safetensors
pip install --pre xformers
使用StableDiffusionPipeline:
(代码部分保持原样,仅翻译注释)
使用长提示词加权管道:
(代码部分保持原样,仅翻译注释)
许可证
PVC v3采用Fair AI Public License 1.0-SD(https://freedevproject.org/faipl-1.0-sd/)发布。如需衍生修改,请遵循许可证要求共享改动。特别感谢ronsor/undeleted(https://undeleted.ronsor.com/)的许可证指导。