Self-QA: Unsupervised Knowledge Guided Language Model Alignment

TL;DR

Self-QA is an unsupervised framework that leverages knowledge to generate large-scale, domain-specific instruction data for language model alignment, reducing reliance on human-annotated datasets.

Contribution

We introduce Self-QA, a novel unsupervised approach that replaces human instruction seeds with knowledge-based data to improve language model instruction tuning.

Findings

Self-QA produces high-quality, domain-specific instruction data.

The method reduces the need for extensive human annotation.

Experiments show improved model alignment with less supervision.

Abstract

Large-scale language models like ChatGPT and GPT-4 have gained attention for their impressive conversational and generative capabilities. However, the creation of supervised paired question-answering data for instruction tuning presents formidable challenges. This endeavor necessitates substantial human effort for data annotation and wrestles with issues concerning data quality, diversity, accuracy, and other related factors. To overcome these obstacles, we introduce an innovative framework named Self-QA, which replaces the traditional practice of human-written instruction seeds with a vast amount of unsupervised knowledge, enabling the model to generate a larger quantity of correct and domain-specific instruction data. The effectiveness of our proposed method is demonstrated through experiments conducted on unsupervised corpora from various domains.