/text-to-sql

An application to write and run SQL queries, returning answers to natural language questions, using langchain and open source LLM models through HuggingFace.

Primary LanguageJupyter NotebookMIT LicenseMIT

Text-to-SQL Copilot

Text-to-SQL Copilot is a tool to support users who see SQL databases as a barrier to actionable insights. Taking your natural language question as input, it uses a generative text model to write a SQL statement based on your data model. Then runs it on your database and analyses the results. And it does this all at no cost using HuggingFace Inference API.

copilot_demo

Setup

Dataset

This was built specifically off of the Spider dataset. Follw these steps to recreate:

  1. Download the data from this Google Drive
  2. Unzip the file
  3. Save the root 'spider' folder under the src/data/raw/ directory

Setup Process

This application pulls the schema information from the SQLite database files and utilizes a locally stored Chroma Vector database to identify which schema to use to answer questions. Run the following commands to compile the database info and build the vector database:

pip3 install -r requirements.txt

python3 setup.py

This will take about 10-15 minutes to fully run.

HuggingFace API Token

Currently, this project relies on the google flan-t5-xxl languauge model. It is accessed for free through the HuggingFace Inference API. In order to use this method, you need to create an API token and save in in a .env file in the root of the repo:

touch .env

Open the .env file and enter your HuggingFace API token:

env_example

Using SQL Copilot

Navigate to the src/app directory and start the program with the following command:

python3 main.py

Then input your question - happy SQL-ing!

Citation

Chase, H. (2022). LangChain [Computer software]. https://github.com/hwchase17/langchain

Yu, T., Zhang, R., Yang, K., Yasunaga, M., Wang, D., Li, Z., ... & Radev, D. (2018). Spider: A large-scale human-labeled dataset for complex and cross-domain semantic parsing and text-to-sql task. arXiv preprint arXiv:1809.08887.