TAPEX

Table Pre-training via Execution

Natural Language ProcessingIntroduced 20003 papers
Source Paper

Description

TAPEX is a conceptually simple and empirically powerful pre-training approach to empower existing models with table reasoning skills. TAPEX realizes table pre-training by learning a neural SQL executor over a synthetic corpus, which is obtained by automatically synthesising executable SQL queries.

Papers Using This Method

QATCH: Benchmarking SQL-centric tasks with Table Representation Learning Models on Your Data2023-09-26TaCube: Pre-computing Data Cubes for Answering Numerical-Reasoning Questions over Tabular Data2022-05-25TAPEX: Table Pre-training via Learning a Neural SQL Executor2021-07-16