tanaos/synthetic-guardrail-dataset-v1

Tanaos-guardrail-v1训练数据集是一个由Tanaos使用Artifex Python库合成创建的数据集，旨在训练和评估防护栏系统，这些系统能够检测、分类或过滤不安全、有害或违反政策的文本内容。该数据集可用于训练审查模型或为聊天机器人、内容生成和面向用户的AI系统整合LLM安全过滤器。数据集包含标记为“安全”或“不安全”的文本样本，其中“不安全”类别包括不当言论过滤、暴力或自伤内容、成人内容、骚扰或欺凌等。此外，数据集还关注隐私保护以及上下文控制，防止机器人收集、暴露或泄露敏感信息，并确保聊天机器人保持在既定目的上。

The Tanaos-guardrail-v1 Training Dataset is synthetically created by Tanaos using the Artifex Python library, designed to train and evaluate guardrail systems that detect, classify, or filter unsafe, harmful, or policy-violating text content. It can be used to train moderation models or integrate LLM safety filters for chatbots, content generation, and user-facing AI systems. The dataset includes text samples labeled as safe or unsafe, with unsafe categories covering profanity, hate speech, violence, adult content, harassment, and more. Additionally, the dataset addresses privacy protection and context control to prevent the bot from collecting, exposing, or leaking sensitive information and to ensure the chatbot stays on its intended purpose.