Datasets › WikiSQL

WikiSQL

Introduced by Victor Zhong et al. in Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning1 Jan 2017 archive 2025-07-28

WikiSQL consists of a corpus of 87,726 hand-annotated SQL query and natural language question pairs. These SQL queries are further split into training (61,297 examples), development (9,145 examples) and test sets (17,284 examples). It can be used for natural language inference tasks related to relational databases.

Source: SQL-to-Text Generation with Graph-to-Sequence Model Image Source: https://blog.einstein.ai/how-to-talk-to-your-database/

Benchmarks archive 2025-07-28

All 4 leaderboards whose dataset resolves to this page shown (sort by any header). "First row" is the archive's own first row at snapshot, in the archive's row order; nothing here re-ranks and metric direction is not asserted.

First row (archive order)PaperCode
Code Generation WikiSQL NL2SQL-RULE Execution Accuracy 89.2 Content Enhanced BERT-based Text-to-SQL Generation guotong1988/NL2SQL-RULE +4 10 Compare
Semantic Parsing WikiSQL NL2SQL-BERT Accuracy 89 Content Enhanced BERT-based Text-to-SQL Generation guotong1988/NL2SQL-RULE +4 5 Compare
Question Answering WikiSQL PieTa Exact Match (EM) 88.55 Piece of Table: A Divide-and-Conquer Approach for... — 2 Compare
SQL-to-Text WikiSQL Graph2Seq-PGE BLEU-4 38.97 Graph2Seq: Graph to Sequence Learning with... IBM/Graph2Seq +3 2 Compare

Papers archive 2025-07-28

15 shown of 15 papers with a leaderboard row on this dataset's benchmarks, newest first. The archive's own "papers using this dataset" list was never published, so this is the benchmark-backed subset; the archive's count for this dataset is 267. The Syntology column is from Syntology's graph (read 2026-09-24), stated per sample; it is not part of any archive number.

DateSamples run Syntology
Piece of Table: A Divide-and-Conquer Approach for Selecting Sub-Tables in Table Question Answering 0 1 10 Dec 2024 not harvested
TabSQLify: Enhancing Reasoning Capabilities of LLMs Through Table Decomposition 2 1 15 Apr 2024 ran 7 of 15 samples (8 unverified; 15 pointer-only for licence)
CABINET: Content Relevance based Noise Reduction for Table Question Answering 1 1 2 Feb 2024 ran 3 of 5 samples (2 unverified; 5 pointer-only for licence)
ReasTAP: Injecting Table Reasoning Skills During Pre-training via Synthetic Reasoning Examples 1 1 22 Oct 2022 ran 0 of 11 samples (11 unverified)
TAPEX: Table Pre-training via Learning a Neural SQL Executor 4 1 16 Jul 2021 not harvested
TAPAS: Weakly Supervised Table Parsing via Pre-training 8 1 5 Apr 2020 ran 0 of 14 samples (14 unverified)
Content Enhanced BERT-based Text-to-SQL Generation 5 2 16 Oct 2019 not harvested
TRANX: A Transition-based Neural Abstract Syntax Parser for Semantic Parsing and Code Generation 4 1 5 Oct 2018 ran 2 of 7 samples (5 unverified; 2 pointer-only for licence)
TypeSQL: Knowledge-based Type-Aware Neural Text-to-SQL Generation 1 2 25 Apr 2018 not harvested
Semantic Parsing with Syntax- and Table-Aware SQL Generation 0 2 23 Apr 2018 not harvested
Graph2Seq: Graph to Sequence Learning with Attention-based Neural Networks 4 1 3 Apr 2018 ran 2 of 10 samples (8 unverified)
Natural Language to Structured Query Generation via Meta-Learning 1 1 2 Mar 2018 not harvested
Bidirectional Attention for SQL Generation 2 1 30 Dec 2017 not harvested
Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning 15 2 31 Aug 2017 not harvested
Gated Graph Sequence Neural Networks 13 1 17 Nov 2015 ran 0 of 12 samples (12 unverified)

Dataset loaders archive 2025-07-28

4 loaders as listed in the archive; links are outbound and not re-checked here.

Tasks archive 2025-07-28

License archive 2025-07-28

BSD 3-Clause License

Modalities archive 2025-07-28

Languages archive 2025-07-28

No language tagged.

Variants archive 2025-07-28

  • WikiSQL

1 variant name, as the archive lists them.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections