QU-NLP at QIAS 2026: Multi-Stage QLoRA Fine-Tuning for Arabic Islamic Inheritance Reasoning

Implement a modular pipeline for high-precision reasoning fine-tuning. This allows domain experts to inject specific rule-based knowledge into LLMs with minimal resource overhead.

Suggested repo: reason-qlora

"Inject complex legal/math reasoning into LLMs on a consumer GPU."

Estimated effort: 30h