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
-
- Easily upload your CSV files for analysis.
- Use the file uploader in the sidebar to upload your CSV file.
-
- 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.
-
- 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:
-
- Visualize your data with scatter plots, bar charts, and heatmaps.
- In the
Data Exploration
tab, visualize your data using various charts:
Table2Text/
│
├── app/
│ ├── init.py
│ ├── model.py
│ └── config.py
│
├── app.py
│
├── README.md
│
└── requirements.txt
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.
This file contains configuration parameters for the model, such as the model ID, maximum output tokens, and the Groq API key.
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.
-
This file lists all the Python dependencies required for the project. Use
pip install -r requirements.txt
to install these dependencies. -
This file can be empty or used to initialize the package.
-
-
- Clone the Repository: Clone the repository to your local machine using Git
git clone https://github.com/KaifAhmad1/Table2Text.git
- Clone the Repository: Clone the repository to your local machine using Git
-
- Navigate to the Project Directory: Navigate to the directory where you cloned the repository
cd Table2Text
- Navigate to the Project Directory: Navigate to the directory where you cloned the repository
-
- Install Dependencies: Install the required Python packages listed in
requirements.txt
using pippip install -r requirements.txt
- Install Dependencies: Install the required Python packages listed in
-
- Run the Streamlit Application: Run the Streamlit application by executing the
app.py
scriptstreamlit run app.py
- Run the Streamlit Application: Run the Streamlit application by executing the
-
- 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.