此模型为画风模型，以黑白水墨笔触为主要素材，因此命名为泼墨

This model is a style transfer model that uses black and white ink strokes as its main material, hence its name "Splatter Ink".

有大模型版本（点击跳转）

With the large model version (click to jump).

注意事项：

Notes:

• 1.CFG范围将会改变风格,但是起码保持在2及以上

• 1. The range of the CFG will change the style, but it should be at least 2 or higher.

2~5: 晕染笔触与黑白对比较多,6及以上 : 墨水变重

2~5: There are more blurred strokes and black and white contrast.6 and above: The ink becomes heavier.

• 2.推荐大模型:2D/2.5D风格模型，真实模型效果不太行（想要追求此类效果的除外）

• 2. Recommended for large models: 2D/2.5D style models, as realistic model effects may not be satisfactory (except for those who want to pursue this type of effect).

• 3.本模型没有触发词,正面提示词（prompts）里推荐添加：simple background,white background,推荐在负面提示词(negative prompts)里增加 chinese text, 减少不明所以的文字出现（但是无法杜绝）

• 3. This model does not have trigger words. Recommended positive prompts include "simple background" and "white background". Recommended negative prompts include "Chinese text" to reduce the appearance of unclear text (but cannot be completely eliminated).

• 4.推荐Lora权重为0.7~1，以下分别为不添加LORA，2D,2.5D,3D主流模型不同权重下的出图效果（提示词为：masterpiece,best quality,1girl,solo）

• 4. Recommended Lora weights are between 0.7-1. The output effects of mainstream models with different weights are shown without LORA, with 2D, 2.5D, and 3D models (prompts include "masterpiece, best quality, 1girl, solo).

以2D模型为底膜，添加角色LORA后，在不同权重下的表现（提示词为：masterpiece,best quality,nahida \(genshin impact\),1girl,white hair,multicolored hair, side ponytail,hair ornament,green eyes,symbol-shaped pupils,pointy ears, jewelry, detached sleeves,bracelet, ,white dress,toeless legwear,swing,sitting,full body,<lora:nahida:1>,）

Using a 2D model as the base, adding the character Nahida's LORA, and its performance at different weights are shown (prompts include "masterpiece, best quality, nahida (genshin impact), 1girl, white hair, multicolored hair, side ponytail, hair ornament, green eyes, symbol-shaped pupils, pointy ears, jewelry, detached sleeves, bracelet, white dress, toeless legwear, swing, sitting, full body, <lora:nahida:1>).

最后欢迎使用后返图，后续可能会视情况迭代更新

Finally, users are welcome to use this model and provide feedback. Iterative updates may be made as needed.

I found a new AI tool Shakker, a best image to image tool. You can try it via https://www.shakker.ai

免责声明：

您基于本模型服务生成的文本内容由您自行维护并对其独立判断后使用，您需在遵守适用法律法规和本服务条款的条件下使用，基于本服务生成内容产生的任何知识产权问题由您自行处理，您需对生成内容负责，本人对由此造成的任何损失不负责任。除法律法规另有规定外，未经本人书面同意，不得将本模型服务集成在任何第三方软件或服务中，亦不得以任何直接间接、明示或暗示方式向其他任何第三方披露生成内容与本服务的关联性，您违反本条约给作者本人所造成的损失有权向您追偿。

Disclaimer:

The text content generated by you based on this model service is maintained by you and used at your own independent judgment. You must use it under the conditions of compliance with applicable laws and regulations and these terms of service. Any intellectual property issues arising from the content generated based on this service are your responsibility. You are responsible for the generated content, and I am not responsible for any losses caused by this. Unless otherwise provided by laws and regulations, without my written consent, this model service cannot be integrated into any third-party software or service. Nor can you disclose the association between the generated content and this service to any third party in any direct or indirect, express or implied manner. If you violate this clause, the author has the right to claim compensation from you for the losses caused to the author.