Skip to content

Latest commit

 

History

History
35 lines (23 loc) · 1.32 KB

README.md

File metadata and controls

35 lines (23 loc) · 1.32 KB
Raw

DTS-SQL

Overview

This repository hosts the code for the DTS-SQL paper, featuring state-of-the-art text-to-SQL models with 7B parameters. Our models demonstrate exceptional performance on the Spider and Spider-SYN datasets, setting new benchmarks in the field.

Repository Contents

Finetuning Scripts

  • schema_linking_generation_finetuning.ipynb: Contains code for finetuning DeepSeek and Mistral models for text-to-SQL tasks.
  • sql_generation_finetuning.ipynb: Dedicated to finetuning processes, specifically focusing on SQL generation.

Inference Scripts

  • schema_linking_generation_inference.ipynb: Script for schema linking generation using the finetuned models.
  • sql_generation_inference.ipynb: Script for SQL generation at inference time using the finetuned models.

Dataset Preparation

  • finetuning_dataset_creator.py: Code for creating the finetuning dataset from the Spider dataset.

Utilities

  • Utils directory: Contains all necessary helper functions for formatting tables and creating the dataset.

Results

  • All of the results are stored in the results directory.

Citation

@article{pourreza2024dts, title={DTS-SQL: Decomposed Text-to-SQL with Small Large Language Models}, author={Pourreza, Mohammadreza and Rafiei, Davood}, journal={arXiv preprint arXiv:2402.01117}, year={2024} }