README.md

Table2Text

Table2Text is a Streamlit application powered by Open Source techstack Mistral-8x7b-32k language model by Mistral AI and the Langchain framework. It assists with queries related to tabular data, such as CSV or Pandas DataFrame, which consist of textual and numerical information.

• This application has three Sections:
  • Data Preview
  • Query
  • Data Exploration
• Platform Link: Table2Text

Features and Usage:

• Upload CSV Files:

  • Easily upload your CSV files for analysis.
  • Use the file uploader in the sidebar to upload your CSV file.

• Data Preview:

  • Preview and filter your data directly in the app.
  • View and filter your data in the Data Preview tab.
  • Select columns and filter values to customize the data displayed.
  • Download the filtered data if needed.

• 

• Natural Language Queries:

  • Ask questions about your data and get instant real-time responses using Mistral AI and Groq.
  • Navigate to the Query tab to ask questions about your data.
  • Enter your query in the text input field and get instant responses.
  • Follow up with additional questions to refine your analysis.

• Query:

  • 

• Follow Up:

  • 

• Data Exploration:

  • Visualize your data with scatter plots, bar charts, and heatmaps.
  • In the Data Exploration tab, visualize your data using various charts:

• Statistical Summary: -

• Scatter Plot: -

• Bar Chart: -

• Box Plots: -

Repository Structure

Table2Text/
│
├── app/
│ ├── init.py 
│ ├── model.py 
│ └── config.py 
│
├── app.py 
│
├── README.md 
│
└── requirements.txt 

Detailed Explanation of Project Files

• app.py

This is the main script that runs the Streamlit application. It sets up the web interface, handles file uploads, and interacts with the chatbot model to process queries.

• config.py

This file contains configuration parameters for the model, such as the model ID, maximum output tokens, and the Groq API key.

• model.py

This script uses Langchain to analyze tabular data with natural language queries. It initializes a language model, creates a vector store and retriever, and sets up memory for chat history. The chain can answer queries about the data, such as average age, correlation, and turnover rate. It also provides examples of good and bad queries and data interpretation instructions.

• requirements.txt

  This file lists all the Python dependencies required for the project. Use pip install -r requirements.txt to install these dependencies.

• __init__.py

  This file can be empty or used to initialize the package.

• How to run on Windows Locally:

  • 1. Clone the Repository: Clone the repository to your local machine using Git

       git clone https://github.com/KaifAhmad1/Table2Text.git

  • 2. Navigate to the Project Directory: Navigate to the directory where you cloned the repository

       cd Table2Text

  • 3. Install Dependencies: Install the required Python packages listed in requirements.txt using pip

       pip install -r requirements.txt

  • 4. Run the Streamlit Application: Run the Streamlit application by executing the app.py script

       streamlit run app.py

Acknowledgements:

• Langchain LLM orchestration Framework for Builinding Application
• Streamlit For Building Simple but Effective UI
• Mistral AI For Open Source Mistral 8*7B LLM
• FAISS For Efficient Similarity Search in Tabular Data.
• Groq For Real-Time Responses powered by Low-Latency LPU Inference.