Skip to content

Security: zararogers338-hash/ForgeX-Lab

Security

SECURITY.md

安全说明 / Security Policy

中文

ForgeX 是本地 AI 训练工具,可能处理模型权重、数据集、API Key 和用户文档。请注意:

  • 不要把隐私数据、商业数据或受版权限制的数据直接提交到仓库。
  • 不要提交 .envlicense.key、API Key、私钥或个人模型缓存。
  • 只加载你信任的模型仓库;部分流程可能使用 trust_remote_code=True
  • 训练和导出流程可能调用外部命令、下载转换脚本或写入大量文件。
  • 使用第三方 API 进行蒸馏时,请遵守服务条款和数据政策。
  • 生成模型可能输出错误、有害或侵权内容;发布前请进行安全评估。

如发现安全问题,请优先私下联系维护者,不要直接公开利用细节。


English

ForgeX is a local AI training tool and may process model weights, datasets, API keys, and user documents. Please note:

  • Do not commit private data, commercial data, or copyrighted datasets.
  • Do not commit .env, license.key, API keys, private keys, or local model caches.
  • Only load model repositories you trust; some paths may use trust_remote_code=True.
  • Training/export workflows may invoke external commands, download conversion scripts, or write large files.
  • When using third-party APIs for distillation, follow their terms and data policies.
  • Generated models may produce inaccurate, harmful, or infringing outputs; evaluate before release.

If you find a security issue, please contact the maintainer privately first instead of publishing exploit details.

There aren't any published security advisories