sallani/EUAIAct-Qwen2.5-0.5B-Edge
The sallani/EUAIAct-Qwen2.5-0.5B-Edge is a 494 million parameter Small Language Model (SLM) developed by sallani, fine-tuned specifically on the EU AI Act (Regulation EU 2024/1689) and GDPR compliance corpus. This compact model is designed for offline, on-device inference across various platforms including mobile browsers via WebGPU, Apple Silicon via MLX, and other machines via GGUF. It specializes in providing guidance and answering questions related to EU AI Act and GDPR compliance in both French and English.
Loading preview...
Overview
EUAIAct-Qwen2.5-0.5B-Edge is a 494 million parameter Small Language Model (SLM) developed by sallani. It is a LoRA fine-tune of the Qwen/Qwen2.5-0.5B-Instruct base model, specifically trained on 300 curated Q&A pairs covering the EU AI Act (Regulation EU 2024/1689) and GDPR compliance. This model is engineered for offline-first, on-device operation, eliminating the need for cloud APIs or internet connectivity after initial download.
Key Capabilities
- Specialized Knowledge: Deep expertise in EU AI Act and GDPR regulations.
- Offline & On-Device: Runs entirely on-device across mobile browsers (WebGPU), Apple Silicon (MLX), and other platforms (GGUF).
- Multilingual Support: Functions in both French and English.
- Lightweight & Efficient: With 494M parameters, it's compact (350 MB GGUF) and optimized for edge deployment.
- Open Weights: Released under Apache 2.0, allowing full auditability and modification.
Intended Use Cases
- EU AI Act Compliance Q&A: Answering questions regarding the regulation.
- GDPR Guidance: Providing insights on GDPR implications for AI systems.
- AI Risk Classification: Assisting with the classification of AI systems.
- GRC Documentation Support: Aiding in governance, risk, and compliance documentation.
This model is intended for advisory use by CISOs, DPOs, GRC consultants, legal counsel, and compliance officers, but it is not a substitute for formal legal advice.