Yee-R1-mini is a 2 billion parameter AI data security expert system developed by 广州熠数信息技术有限公司 (Guangzhou Shining Data Information Technology Co., Ltd.). Built on the Qwen3-1.7B architecture with a 40960 token context length, it integrates data classification, security auditing, and protection capabilities. This model is specialized for providing lightweight, intelligent data security solutions across industrial, government, and telecommunications sectors, focusing on compliance, visibility, control, and prevention.
Loading preview...
What is Yee-R1-mini?
Yee-R1-mini is an AI data security expert system developed by 广州熠数信息技术有限公司 (Guangzhou Shining Data Information Technology Co., Ltd.). It is built upon the Qwen3-1.7B large language model architecture and is specifically designed to address data security challenges across various industries, including industrial, government, and telecommunications.
Key Capabilities
- Specialized Data Security Focus: Integrates capabilities for data classification, security auditing, and leakage prevention across cloud, network, and endpoint scenarios.
- Dual-Mode Reasoning: Features a unique "Thinking Mode" for complex logical tasks (e.g., code analysis, mathematical computation, strategy formulation) and a "Non-Thinking Mode" for faster responses in daily conversations.
- Agentic Functionality: Leverages the Qwen-Agent framework to enable tool calling, allowing interaction with external systems like databases, log analyzers, and APIs for automated task execution.
- High Compatibility: Supports diverse deployment methods including local execution, Docker, Kubernetes, and SaaS APIs, and is compatible with major inference frameworks such as HuggingFace Transformers, vLLM, SGLang, and Ollama.
- Performance: Achieves a comprehensive score of 77.48 in CS-Eval benchmarks across various security domains, demonstrating proficiency in areas like AI & Network Security (84.65) and Supply Chain Security (86.71).
When to Use Yee-R1-mini
This model is ideal for organizations seeking an intelligent, lightweight solution for data security management, particularly for tasks requiring deep analysis, policy enforcement, and automated responses within a data security context. Its dual-mode reasoning and agent capabilities make it suitable for both complex security operations and routine inquiries.