This is a fill lora. Use as same as fill workflow and load this lora after loading fill model

版权声明与使用协议（研究限定）

本声明适用于所有通过本LoRA扩展模型（以下简称"本模型"）生成的图像内容：

1. 版权归属原则

   • 用户使用本模型生成的所有图像，其版权、知识产权及相关法律责任均由生成者独立承担

   • 生成内容与本模型开发者、模型提供者及训练数据源无任何关联

2. 使用限制与免责条款

   • 本模型仅限于研究及学术用途，禁止任何形式的商业化应用与传播

   • 本模型遵循"按现状"原则提供，不承诺生成内容的合法性、安全性或适用性

   • 禁止将本模型用于任何违反所在地法律法规的用途，包括但不限于：制作虚假信息、侵犯他人权益、生成非法内容等

3. 用户责任条款

   • 使用者须对生成内容进行合法性审查并承担全部责任

   • 因使用本模型产生的任何纠纷或损失，开发者及模型提供方不承担连带责任

4. 协议变更说明

   • 本声明最终解释权归模型开发者所有

   • 开发者保留随时更新本声明的权利，恕不另行通知

Copyright Notice & Usage Agreement (Research Only)

This statement applies to all images generated through this LoRA extension model (hereinafter referred to as "the Model"):

1. Copyright Ownership

   • All copyrights, intellectual property rights, and legal responsibilities for images generated using this Model shall be solely borne by the user

   • Generated content has no affiliation with the Model developer, provider, or training data sources

2. Usage Restrictions & Disclaimer

   • The Model is strictly limited to research and academic purposes, commercial use and distribution are expressly prohibited

   • The Model is provided "AS IS" without warranties of legality, safety, or fitness for purpose

   • Any illegal use violating local laws/regulations is strictly prohibited, including but not limited to: creating false information, infringing rights, generating unlawful content, etc.

3. User Responsibilities

   • Users must conduct legal compliance checks on generated content and assume full responsibility

   • The developer/model provider bears no liability for disputes or damages arising from Model usage

4. Agreement Modifications

   • Final interpretation rights belong to the Model developer

   • The developer reserves the right to update this notice without prior notice

This version is the alpha version. The objective of this lora is to improve state transfer learning script of flux fill or any dit model.

This lora maintains good enviromental shadow and minimize the side effect.

This alpha lora is still has some bad case or weak case which is not able to do its objective but the lora training would be stopped as this version.

📬 Contact

𝕏: [@Lrzjason](https://twitter.com/Lrzjason)

📧: [[email protected]](mailto:[email protected])

QQ Group: 866612947

WeChat ID: fkdeai

CivitAI: [xiaozhijason](https://civarchive.com/user/xiaozhijason)

Sponsors me for more open source projects: