chatmydata

interact with your data in natural language, powered by LLM. automatically generate visualization in pie chart, bar chart, and others. Components used:

• Google text-bison@001 model from Vertex AI Palm API
• Langchain
• Streamlit

Credit:

• langchain_csv
• chartgpt

Use case

• upload sample csv : bankfinance.csv
• ask questions about data inside csv
• depending on the prompt, chatmydata will show the response in table, chart, or text

Installation

Tested on Python 3.9x. Create Python virtual enviroment and install dependencies:

pip install -r requirements.txt

Run chatmydata

streamlit run chatmydata.py

Usage examples

chatmydata was tested to receive questions in English and Bahasa Indonesia.

Example 1:

• Ask in English

Example 2:

Example 3:

Example 4:

• Now switch question to Bahasa Indonesia

Example 5:

Run as Docker Image

Use provided Dockerfile to build a docker image and run chatmydata on container. Container is configured to run on port 8080. Edit Docker file to change port and entry point eccordingly.

1. Build docker image

   docker build -t your-image-name .
   

2. Run the image locally

   docker run -p 8080:8080 your-image-name