taoyds/spider resource

Health v2 · maintenance only

Flags: no_releases

Adoption not part of the score

1097 stars · 227 forks observed · 2026-08-28

What it is AI-extracted, prompt v1, taxonomy v1, 2026-08-30, confidence not recorded

Spider is a large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-SQL, with 10,181 questions and 5,693 SQL queries across 200 databases. This repository contains the evaluation scripts, preprocessing code, and baseline models from the EMNLP 2018 paper.

Use cases

• train and evaluate text-to-SQL models
• benchmark semantic parsing systems on cross-domain databases
• evaluate generated SQL queries against gold labels
• build natural language interfaces for relational databases
• compare baseline text-to-SQL model performance
• research generalization to unseen database schemas

When to choose

• you need a standard benchmark for text-to-SQL or semantic parsing research
• you want official evaluation scripts for the Spider challenge leaderboard
• you need a large cross-domain dataset of natural language questions paired with SQL queries

When to avoid

• you need a production text-to-SQL system rather than a research dataset
• you want a maintained end-to-end model rather than baselines and evaluation code
• you need a simpler single-domain text-to-SQL dataset

Facets

dataset · maturity stable

nlp machine-learning parser databases machine-learning artificial-intelligence python cross-platform text-to-sql semantic-parsing benchmark dataset evaluation-scripts baselines evaluation natural-language-processing

Member repositories

For agents

markdown · JSON · MCP: product_card(name="taoyds/spider")