cycloevan/gdpr_gemma-2-2b

TEXT GENERATIONConcurrency Cost:1Model Size:2.6BQuant:BF16Ctx Length:8kPublished:Sep 29, 2024License:apache-2.0Architecture:Transformer0.0K Open Weights Cold

GDPR-Gemma-2-2B by cycloevan is a 2.6 billion parameter instruction-tuned model, fine-tuned from Google's Gemma-2-2B-IT, specifically for English General Data Protection Regulation (GDPR) Q&A. It utilizes a unique 3-stage alignment pipeline (SFT, Dynamic Rejection Sampling, DPO) to enhance legal correctness and compliance alignment. This model excels at providing informational guidance on GDPR articles and principles, making it suitable for educational explanations, drafting compliance summaries, and internal training materials, with a context length of 8192 tokens.

Loading preview...

GDPR-Gemma-2-2B: Specialized GDPR Compliance Assistant

GDPR-Gemma-2-2B, developed by seok-hee97 (cycloevan), is a 2.6 billion parameter model fine-tuned from google/gemma-2-2b-it specifically for English GDPR (General Data Protection Regulation) Q&A. It stands out due to its 3-stage alignment pipeline: Supervised Fine-Tuning (SFT) for domain knowledge, Dynamic Rejection Sampling to create realistic preference pairs, and Direct Preference Optimization (DPO) for aligning with expert answers. This resource-friendly training uses QLoRA.

Key Capabilities

  • Enhanced Legal Accuracy: DPO significantly improves legal correctness, GDPR article citation accuracy, and compliance alignment compared to its base and SFT-only variants.
  • Specialized Knowledge: Provides detailed explanations of GDPR articles and principles.
  • Informational Guidance: Designed for educational purposes, drafting first-pass compliance summaries, and internal training.
  • Resource-Efficient Training: Utilizes QLoRA for fine-tuning, making it accessible.

Good for

  • Educational explanations of GDPR articles and principles.
  • Drafting initial compliance summaries and internal training materials.
  • Prototyping GDPR-aware chatbots.
  • Decision-support tools in GDPR contexts (with verification).

Disclaimer: This model offers informational guidance only and does not constitute legal advice. Always consult a qualified legal professional for binding GDPR compliance decisions.

Finetunes
1 models