html-table-question-answering

Baseline: hyunwoongko/kobart
Fine-Tuning: AI-Hub: 표 정보 질의응답 데이터

Introduction

This project is practice code for 한국소프트웨어종합학술대회(KSC) 2024 표 이해를 위한 특수 토큰을 활용한 HTML 표 질의응답 모델 [Paper].

Reference

BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation
Know What You Don't Know: Unanswerable Questions for SQuAD
Chain-of-Table: Evolving Tables in the Reasoning Chain for Table Understanding
국립국어원 모두의말뭉치