This project is a Retrieval-Augmented Generation (RAG) System that allows users to upload PDF files, process their contents, and interact with them through conversational queries. The app leverages Groq's Mixtral-8x7b LLM model (DeepSeek-R1 variant) for advanced natural language understanding and answering questions with high accuracy. The backend features FAISS for vector search and OpenAI embeddings for efficient similarity-based retrieval.
- PDF Upload and Processing: Upload multiple PDF files, extract their text, and store it in a searchable FAISS vector database.
- Conversational Querying: Ask natural language questions about the content of the uploaded PDFs.
- Accurate Answers: The app uses the DeepSeek-R1 model from Groq, which provides detailed answers, or informs when the information is not available in the context.
- RAG Workflow: Combines information retrieval with generative capabilities for seamless and reliable responses.
- Streamlit Interface: A user-friendly interface for uploading files, processing text, and interacting with the system.
- LangChain: Framework for building LLM-powered workflows.
- FAISS: Scalable vector database for similarity search.
- OpenAI Embeddings: Used for creating embeddings from the extracted text.
- Groq LLM: High-performing language model for answering questions.
- Streamlit: Simplified app creation for interactive dashboards and interfaces.
- Environment management via
dotenv.
Ensure you have the following installed:
- Python 3.9+
- Pip
- A valid Groq API key
-
Clone the repository:
git clone <repository-url> cd <repository-folder>
-
Install dependencies:
pip install -r requirements.txt
-
Set up environment variables:
- Create a
.envfile in the root directory. - Add your Groq API key:
GROQ_API_KEY=your-groq-api-key
- Create a
-
Run the application:
streamlit run app.py
- Use the sidebar to upload one or more PDF files.
- Click on Submit & Process to extract text and create a searchable vector database.
- Wait for the success message indicating the PDFs have been indexed.
- Enter your question in the text input field on the main page.
- The app will retrieve relevant information from the vector database and generate an accurate response.
.
├── app.py # Main Streamlit application
├── requirements.txt # Required Python libraries
├── .env # Environment variables (Groq API key)
├── faiss_index/ # Folder containing the FAISS vector store
└── README.md # Project documentation
Upload research papers in PDF format and ask detailed questions about their content.
Example Question: "What are the main findings of the study on page 3?"
Upload contracts or legal documents to extract critical clauses or terms.
Example Question: "What is the termination clause in this agreement?"
Upload textbooks or study material to quickly find answers to specific topics.
Example Question: "Explain the concept of retrieval-augmented generation."
- Support for Additional File Types: Include support for Word, Excel, or plain text files.
- Customizable Chunk Size: Allow users to customize chunking parameters for optimized retrieval.
- Cloud Integration: Enable cloud storage and retrieval for larger datasets.
- Multimodal Querying: Add support for images and diagrams in PDFs.
- Groq: For their Mixtral-8x7b model, which powers the app's conversational capabilities.
- LangChain: For providing a robust framework for integrating LLMs and vector databases.
- FAISS: For efficient and scalable similarity search.
This project is licensed under the MIT License.
For questions or feedback, please reach out to:
- Name: Siddharth Kharche
- Email: siddukharche04@gmail.com
- GitHub: siddharth-Kharche
Happy Conversing! 📚 💬