An End-to-End Ukrainian RAG for Local Deployment. Optimized Hybrid Search and Lightweight Generation

arXiv cs.CL / 4/27/2026

📰 NewsDeveloper Stack & InfrastructureTools & Practical UsageModels & Research

共有:

Key Points

The paper introduces an end-to-end, locally deployable Ukrainian Retrieval-Augmented Generation (RAG) system for document question answering that secured 2nd place in the UNLP 2026 Shared Task.
It uses a custom two-stage hybrid search pipeline to retrieve relevant document pages, then generates grounded answers using a Ukrainian language model fine-tuned on synthetic data.
The authors compress the model to enable lightweight deployment, aiming to maintain answer quality while reducing compute requirements.
Experiments under strict computational limits show that verifiable, high-quality AI QA can run locally on resource-constrained hardware without sacrificing accuracy.

Abstract

This paper presents a highly efficient Retrieval-Augmented Generation (RAG) system built specifically for Ukrainian document question answering, which achieved 2nd place in the UNLP 2026 Shared Task. Our solution features a custom two-stage search pipeline that retrieves relevant document pages, paired with a specialized Ukrainian language model fine-tuned on synthetic data to generate accurate, grounded answers. Finally, we compress the model for lightweight deployment. Evaluated under strict computational limits, our architecture demonstrates that high-quality, verifiable AI question answering can be achieved locally on resource-constrained hardware without sacrificing accuracy.