🚀 Protogen v2.2（动漫）官方版本
Protogen v2.2是一款基于Stable Diffusion的模型，它在大量新的和热门的数据集上进行了微调，能够生成高质量的图像，尤其在动漫风格图像生成上表现出色。同时，它支持使用特定提示词来控制生成效果，并且可以与Dreambooth结合使用。
🚀 快速开始
安装
要运行此模型，请下载model.ckpt或model.safetensor文件，并将其安装到stable-diffusion-webui\models\Stable-diffusion目录中。
运行示例
此模型可以像其他Stable Diffusion模型一样使用。以下是一个使用Python代码生成图像的示例：
from diffusers import StableDiffusionPipeline, DPMSolverMultistepScheduler
import torch
prompt = (
"modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, "
"english medieval witch, black silk vale, pale skin, black silk robe, black cat, necromancy magic, medieval era, "
"photorealistic painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, "
"trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski"
)
model_id = "darkstorm2150/Protogen_v2.2_Official_Release"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda")
image = pipe(prompt, num_inference_steps=25).images[0]
image.save("./result.jpg")
✨ 主要特性
通用信息
Protogen以Stable Diffusion v1 - 5为基础进行预热启动，并在civitai.com上大量新的和热门的大型数据集上进行了微调。
你可以通过在提示词中使用“modelshoot style”来强制使用相机捕捉风格。此外，该模型非常适合使用Dreambooth，只需少量步骤就能生成高保真度的人脸（详见dreambooth）。
粒度自适应学习
粒度自适应学习是一种机器学习技术，它专注于在细粒度级别上调整学习过程，而不是对模型进行全局调整。这种方法使模型能够适应数据中的特定模式或特征，而不是基于一般趋势进行假设。
粒度自适应学习可以通过主动学习等技术实现，主动学习允许模型选择它想要学习的数据；也可以通过强化学习实现，在强化学习中，模型会根据其性能获得反馈并进行调整。此外，还可以通过在线学习等技术实现，在线学习使模型在接收更多数据时自我调整。
粒度自适应学习通常用于数据高度多样化或非平稳的情况，以及模型需要快速适应变化模式的场景，如机器人技术、金融市场和自然语言处理等动态环境。
触发词
modelshoot style
此外，hassan1.4和f222也有可用的触发词，你可能需要通过谷歌搜索来获取它们。
📦 安装指南
要运行此模型，请下载以下文件之一：
然后将下载的文件安装到stable-diffusion-webui\models\Stable-diffusion目录中。
💻 使用示例
基础用法
from diffusers import StableDiffusionPipeline, DPMSolverMultistepScheduler
import torch
prompt = (
"modelshoot style, (extremely detailed CG unity 8k wallpaper), full shot body photo of the most beautiful artwork in the world, "
"english medieval witch, black silk vale, pale skin, black silk robe, black cat, necromancy magic, medieval era, "
"photorealistic painting by Ed Blinkey, Atey Ghailan, Studio Ghibli, by Jeremy Mann, Greg Manchess, Antonio Moro, trending on ArtStation, "
"trending on CGSociety, Intricate, High Detail, Sharp focus, dramatic, photorealistic painting art by midjourney and greg rutkowski"
)
model_id = "darkstorm2150/Protogen_v2.2_Official_Release"
pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)
pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda")
image = pipe(prompt, num_inference_steps=25).images[0]
image.save("./result.jpg")
📚 详细文档
空间支持
我们支持使用Gradio Web UI来运行dreamlike - diffusion - 1.0：
检查点合并数据参考
| 模型 |
Protogen v2.2（动漫） |
Protogen x3.4（照片） |
Protogen x5.3（照片） |
Protogen x5.8（科幻/动漫） |
Protogen x5.9（龙） |
Protogen x7.4（日食） |
Protogen x8.0（新星） |
Protogen x8.6（无限） |
| seek_art_mega v1 |
52.50% |
42.76% |
42.63% |
|
|
|
25.21% |
14.83% |
| modelshoot v1 |
30.00% |
24.44% |
24.37% |
2.56% |
2.05% |
3.48% |
22.91% |
13.48% |
| elldreth v1 |
12.64% |
10.30% |
10.23% |
|
|
|
6.06% |
3.57% |
| photoreal v2 |
|
|
10.00% |
48.64% |
38.91% |
66.33% |
20.49% |
12.06% |
| analogdiffusion v1 |
|
4.75% |
4.50% |
|
|
|
1.75% |
1.03% |
| openjourney v2 |
|
4.51% |
4.28% |
|
|
4.75% |
2.26% |
1.33% |
| hassan1.4 |
2.63% |
2.14% |
2.13% |
|
|
|
1.26% |
0.74% |
| f222 |
2.23% |
1.82% |
1.81% |
|
|
|
1.07% |
0.63% |
| hasdx |
|
|
|
20.00% |
16.00% |
4.07% |
5.01% |
2.95% |
| moistmix |
|
|
|
16.00% |
12.80% |
3.86% |
4.08% |
2.40% |
| roboDiffusion v1 |
|
4.29% |
|
12.80% |
10.24% |
3.67% |
4.41% |
2.60% |
| RPG v3 |
|
5.00% |
|
|
20.00% |
4.29% |
4.29% |
2.52% |
| anything&everything |
|
|
|
|
|
4.51% |
0.56% |
0.33% |
| dreamlikediff v1 |
|
|
|
|
|
5.0% |
0.63% |
0.37% |
| sci - fidiff v1 |
|
|
|
|
|
|
|
3.10% |
| synthwavepunk v2 |
|
|
|
|
|
|
|
3.26% |
| mashupv2 |
|
|
|
|
|
|
|
11.51% |
| dreamshaper 252 |
|
|
|
|
|
|
|
4.04% |
| comicdiff v2 |
|
|
|
|
|
|
|
4.25% |
| artEros |
|
|
|
|
|
|
|
15.00% |
📄 许可证
通过下载此模型，你同意遵守以下许可证的条款：
%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
Transformers 支持多种语言