README.md

metadata

dataset_info:
  features:
    - name: answer
      dtype: string
    - name: question
      dtype: string
    - name: context
      dtype: string
  splits:
    - name: train
      num_bytes: 90244824.04555836
      num_examples: 118695
    - name: test
      num_bytes: 11154652.165005254
      num_examples: 14835
    - name: validation
      num_bytes: 11267107.641870596
      num_examples: 14838
  download_size: 24508769
  dataset_size: 112666583.85243422
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
      - split: test
        path: data/test-*
      - split: validation
        path: data/validation-*
task_categories:
  - question-answering
  - zero-shot-classification
  - text2text-generation
language:
  - en
size_categories:
  - 100K<n<1M

Overview

This dataset is build from b-mc2/sql-create-context, Clinton/Text-to-sql-v1 and knowrohit07/know_sql.

A total of 148368 examples consist of natural language queries, SQL CREATE TABLE statements, and SQL Query answering the question using the CREATE statement as context.

Next Step

• Further augment the data by converting queries and CREATE TABLE statements into different SQL dialects.
• Support other informative contexts beyond CREATE TABLE.