tilake/China-Chic-illustration
DreamBooth model for China-Chic-illustration
This is the first model for China-Chic illustration (国潮插画) style painting. The model is based on Stable Diffusion model, fine-tuned the China-Chic illustration style taught to Stable Diffusion with DreamBooth. It is trained by tilake AIGC group on the own dataset. It can be used by modifying the instance_prompt: style of \<guo-chao\> illustration
全球首个 “国潮插画” 风格定制化模型发布。该模型基于 Stable Diffusion model,并采用 DreamBooth 微调。该模型由 TiLake AIGC Group 在内部数据集上训练而来。 你可以通过扩充以下 instance_prompt 来使用国潮插画模型绘制各种画作: style of \<guo-chao\> illustration
Gradio
We support a Gradio Web UI to run China-Chic-illustration: 
Description
This is a Stable Diffusion model fine-tuned on China-Chic illustration style images. It is trained by Tilake AIGC Group. If you like this model, click the \[❤ like\] button!
国潮插画是将传统文化和现代潮流审美进行结合的一种插画形式,开放该模型旨在给相关艺术家和工作者提供灵感和创作思路。如果喜欢该模型,欢迎点亮网页最上方的【like】按钮~
Examples
- Prompt: ``
a cute rabbit in red clothes, in the style of <guo-chao> illustration, trending on artstation, masterpiece, best quality``(国潮兔) <img width="200px" height="200px" src="https://huggingface.co/tilake/China-Chic-illustration/resolve/main/example/1.jpg"> - Prompt: ``
dragon dance, in the style of <guo-chao> illustration, trending on artstation, masterpiece, best quality``(国潮舞龙) <img width="200px" height="200px" src="https://huggingface.co/tilake/China-Chic-illustration/resolve/main/example/2.jpg"> - Prompt: ``
fireworks, in the style of <guo-chao> illustration, trending on artstation, masterpiece, best quality``(国潮烟花) <img width="200px" height="200px" src="https://huggingface.co/tilake/China-Chic-illustration/resolve/main/example/3.jpg"> - Prompt: ``
a snowman, fireworks in the background, in the style of <guo-chao> illustration, trending on artstation, masterpiece, best quality``(国潮雪人) <img width="200px" height="200px" src="https://huggingface.co/tilake/China-Chic-illustration/resolve/main/example/4.jpg">
Usage
from diffusers import StableDiffusionPipeline
# guidance_scale=8.8 may be the best
pipeline = StableDiffusionPipeline.from_pretrained('tilake/China-Chic-illustration')
image = pipeline("style of <guo-chao> illustration, a rabbit wearing sunglasses").images[0]
image