应用于欧美休闲游戏人物原画

模型使用须知！

一定要记得加触发词：oumeikatong1

模型权重在0.75到0.85之间，0.8最佳

注意提示词引导系数CFG： 3.5~5

提示词引导系数CFG为 3.5~5

效果最好：

叠加欧美卡通休闲游戏风格V1LORA使用

大模型：F.1-dev-fp8或者F.1-dev-fp16

推荐迭代步数：35步

采样方法：Euler

高清修复：放大算法R-ESRGAN 4x+ Anime6B

放大倍数1.5~2之间

推荐宽高比：1024x1024

在线使用 ， 如需源文件可+V jingxiang011101

免责声明

本人在本网站发布的所有AI生成图片，仅供个人学习和交流使用，非会员模型不得用于任何商业用途，包括但不限于二次销售转载等。

任何个人或组织不得将本人发布的模型生成图片用于训练、微调模型发布在任何网站上，无论是开源、闭源、商业或非商业用途。禁止二次上传或发布模型生成的图片至其他网站、社交媒体、应用程序或其他任何平台，未经授权不得进行任何形式的传播。

Used for character concept art in European and American casual games

Model Usage Instructions!

Remember to add the trigger word: oumeikatong1

Model weight should be between 0.75 and 0.85, with 0.8 being optimal.

Note: The prompt word guidance coefficient (CFG) should be between 3.5 and 5.

The prompt word guidance coefficient (CFG) should be between 3.5 and 5.

Best results:

Use with V1LORA overlays in a European and American cartoon casual game style.

Large model: F.1-dev-fp8 or F.1-dev-fp16

Recommended number of iterations: 35

Sampling method: Euler

HD restoration: Upscaling algorithm R-ESRGAN 4x + Anime6B

Upscaling factor between 1.5 and 2

Recommended aspect ratio: 1024x1024

For online use, please add V jingxiang011101 if you need the source file.

Disclaimer

All AI-generated images posted on this website are for personal learning and communication only. Non-member models may not be used for any commercial purpose, including but not limited to secondary sales and reprinting.

No individual or organization may use the model-generated images I publish for training or fine-tuning models on any website, whether open source, closed source, commercial, or non-commercial. Re-uploading or publishing model-generated images to other websites, social media, applications, or any other platforms is prohibited, and any form of unauthorized dissemination is prohibited.