datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
RBAC-Text2SQL-Benchmark
RBAC-Text2SQL Benchmark
Role-conditioned Text-to-SQL instances for evaluating whether LLMs generate SQL that
respects Role-Based Access Control (RBAC) constraints. Each instance pairs a natural
language question with a role policy; the model must either produce a correct SQL query
that touches only authorized resources, or refuse with Sorry, I cannot answer.
Code, evaluation harness, and reproduction instructions:
https://github.com/2020dfff/RBAC-Text2SQL-Benchmark… See the full description on the dataset page: https://huggingface.co/datasets/sharkiefff/RBAC-Text2SQL-Benchmark.duckdb-text2sql-25k
Dataset Summary
The duckdb-text2sql-25k dataset contains 25,000 DuckDB text-2-sql pairs covering diverse aspects of DuckDB's SQL syntax.
We synthesized this dataset using Mixtral 8x7B, based on DuckDB's v0.9.2 documentation and Spider schemas that were translated to DuckDB syntax and enriched with nested type columns.
Each training sample consists of a natural language prompt, a corresponding (optional) schema, and a resulting query. Each pair furthermore has a category property… See the full description on the dataset page: https://huggingface.co/datasets/motherduckdb/duckdb-text2sql-25k.chichewa-text2sql
Chichewa Text-to-SQL
The first structured Text-to-SQL benchmark for Chichewa, a low-resource Bantu language spoken by over 12 million people in Malawi and neighboring regions.
The dataset contains 400 manually curated natural language–SQL pairs in both Chichewa (Nyanja) and English, grounded in a unified relational SQLite database covering five real-world domains from Malawi.
Dataset Summary
This benchmark was constructed to investigate the adaptation of Large… See the full description on the dataset page: https://huggingface.co/datasets/johneze/chichewa-text2sql.Text2SQL-Decisions
Text2SQL-Decisions v0.3
28,081 English-only examples across two separate configurations. Both are template-generated, execution-validated drafts, not human-reviewed benchmarks. No model has been fine-tuned as part of this release.
Configuration
Rows
Task
License
default
25,000
Four-candidate SQL plan selection on public sensor and e-commerce data
CC BY 4.0
meter
3,081
Bounded meter planning decisions, including conversational context
CC0-1.0
The configurations… See the full description on the dataset page: https://huggingface.co/datasets/Chulinz/Text2SQL-Decisions.spider-text-2-sqlText2SQL-Decisions-Benchmark
Text2SQL-Decisions Benchmark
English evaluations for choosing SQL plans and answering database questions. This repository contains a 24-question synthetic test, 20 application-level questions, and model results on the separate Text2SQL-Decisions dataset.
Model results
Each model received the same question, database schema, and four SQL candidates. Reference answers were excluded from model input. Accuracy measures selection of the correct candidate.
Model… See the full description on the dataset page: https://huggingface.co/datasets/Chulinz/Text2SQL-Decisions-Benchmark.Text2SQLtext2sql-oracle-postgres
Oracle / PostgreSQL text-to-SQL
Instruction data for fine-tuning google/gemma-3-270m-it (or any chat model) to emit a single dialect-correct SQL statement.
804 rows, 402 Oracle / 402 PostgreSQL
7 schemas: hr, sales, banking, inventory, tickets, university, logistics
Splits: 684 / 60 / 60 (grouped so paraphrases of the same SQL stay in one split)
Load
from datasets import load_dataset
ds = load_dataset("chabab/text2sql-oracle-postgres")
Record… See the full description on the dataset page: https://huggingface.co/datasets/chabab/text2sql-oracle-postgres.text2sqltext2sql-omnisql-style
Dialect: SQLite
Dataset
Source Paper
Samples Used
Notes
Links
Spider
Spider: A Large-Scale Human-Labeled Dataset...
7,000
Seluruh training split digunakan.
Link Google Drive Donwload
BIRD23-Train-Filtered
A BIg Bench for Large-Scale Database Grounded Text-to-SQLs
6,626
Menggunakan subset bird23-train-filtered.
HuggingFace Dataset
SynSQL-2.5M (Filtered)
OmniSQL: Synthesizing High-quality Text-to-SQL Data at Scale
7,000
Hasil filtering berdasarkan question style dan… See the full description on the dataset page: https://huggingface.co/datasets/NafishZaldinanda/text2sql-omnisql-style.text2sql-dataset
Overview
This dataset builds from WikiSQL and Spider.
There are 78,577 examples of natural language queries, SQL CREATE TABLE statements, and SQL Query answering the question using the CREATE statement as context. This dataset was built with text-to-sql LLMs in mind, intending to prevent hallucination of column and table names often seen when trained on text-to-sql datasets. The CREATE TABLE statement can often be copy and pasted from different DBMS and provides table names, column… See the full description on the dataset page: https://huggingface.co/datasets/dipanjanS/text2sql-dataset.Text2SQLpostGIS_text2sqlflock-off-s1-text-2-sqltext2sql
Text2SQL SFT Dataset
This dataset is a combined and cleaned version of Spider, BIRD, and Gretel datasets, formatted specifically for Instruction Fine-Tuning of Small Language Models (SLMs).
Dataset Structure
Each record contains:
instruction: The task description.
input: The Database Schema and the Natural Language Question.
output: The canonicalized SQL query.
metadata: A dictionary containing dataset source and db_id.
Usage
from datasets import load_dataset… See the full description on the dataset page: https://huggingface.co/datasets/rupal02/text2sql.yasserrmd__Text2SQL-1.5B-details
Dataset Card for Evaluation run of yasserrmd/Text2SQL-1.5B
Dataset automatically created during the evaluation run of model yasserrmd/Text2SQL-1.5B
The dataset is composed of 38 configuration(s), each one corresponding to one of the evaluated task.
The dataset has been created from 1 run(s). Each run can be found as a specific split in each configuration, the split being named using the timestamp of the run.The "train" split is always pointing to the latest results.
An additional… See the full description on the dataset page: https://huggingface.co/datasets/open-llm-leaderboard/yasserrmd__Text2SQL-1.5B-details.Text2SQL-GPTtext2sql_challegeMIRDC_TEXT2SQLtext2sql
Text2SQL SFT Dataset
This dataset is a combined and cleaned version of Spider, BIRD, and Gretel datasets, formatted specifically for Instruction Fine-Tuning of Small Language Models (SLMs).
Dataset Structure
Each record contains:
instruction: The task description.
input: The Database Schema and the Natural Language Question.
output: The canonicalized SQL query.
metadata: A dictionary containing dataset source and db_id.
Usage
from datasets import load_dataset… See the full description on the dataset page: https://huggingface.co/datasets/anuj6316/text2sql.text2sql_extended_datasettext2sqlLlama-Text2SQLThis is an adjusted column names version from DominikLindorfer/SQL-LLaMA (https://huggingface.co/datasets/DominikLindorfer/SQL-LLaMA)
text2sql_step2text2sqlmcut_text2sqlflock-off-s1-text-2-sqlbird-text2sql-bench-zhText2SQLtext_2_sql