Question 1

What are SQL-to-Text datasets used for?

Accepted Answer

SQL-to-Text datasets are collections of labelled or raw data used to train, fine-tune, and evaluate models on the sql-to-text task. This page lists 3 such datasets, each linking to its source and paper.

Question 2

Which SQL-to-Text dataset is best for benchmarking?

Accepted Answer

1 of these SQL-to-Text datasets are tracked as benchmarks, including Semantic Parsing in Context (SParC). See the Benchmarks section for model leaderboards.

Question 3

How many SQL-to-Text datasets are there?

Accepted Answer

We catalog 3 SQL-to-Text datasets in one searchable directory.

SQL-to-Text Datasets

What languages do sql-to-text datasets cover?

Explore other dataset tasks

Frequently asked questions