Z-Image-Turbo-Art-GGUF

Z-Image-Turbo-Art with various GGUF versions.

Z-Image-Turbo-Art

本模型为 Z-Image 与 Z-Image-Turbo 的融合微调模型，抽取了部分 Z-Image 基座模型的风格化能力，与 Z-Image-Turbo 分层融合后再进行快速微调而成，<u>这目前还只是如何充分利用 Z-Image 基座模型方面的一次尝试</u>。相比官方原 Turbo 模型，本模型图像更清晰、风格化能力更强，但模型在人像，特别是皮肤的细腻度方面有所下降，文字渲染能力基本保持。具体效果参见如下示例图片，本模型易于使用、参数适配性和LoRA兼容性良好，个人认为本模型更适合艺术风格与广告装饰类图片的生成与创作。

This model is a fine-tuned fusion of Z-Image and Z-Image-Turbo. It extracts some of the stylization capabilities from the Z-Image base model and then performs a layered fusion with Z-Image-Turbo followed by quick fine-tuning, <u>This is just an attempt to fully utilize the Z-Image base model currently</u>. Compared to the official models, this model images are clearer and the stylization capability is stronger, but the model has reduced delicacy in portraits, especially on skin, while text rendering capability is largely maintained. The specific effects can be seen in the sample images below. This model is easy to use, with good parameter adaptability and LoRA compatibility. Personally, I think this model is more suitable for generating and creating images in artistic styles and advertising decoration.

使用方法：

basic workflow pls refer to: Z-Image-Turbo offical workflow

inference cfg: 1.0；

inference steps: 8 -12；

sampler / scheduler: ddim / simple ，or any other。

Also on: Modelscope.cn, Civitai

模型许可：

请遵循 Z-Image 模型的 Apache-2.0 许可。

Please follow the Apache-2.0 open source license for the Z-Image model.