Skip to content

Latest commit

 

History

History
73 lines (48 loc) · 2.91 KB

README.md

File metadata and controls

73 lines (48 loc) · 2.91 KB
Raw

PDFGemini_Bot

Project Overview

PDFGem Chat is an interactive chat interface designed for querying information from uploaded PDF files. This project utilizes Streamlit, PyPDF2, LangChain, Google Generative AI, and FAISS to create a seamless experience for users to ask questions related to the content of PDF documents.

image

Components

1. User Interface

  • Developed using the Streamlit library for a user-friendly experience.
  • Users can ask questions about the content of uploaded PDF files.

2. PDF Processing

  • Extracts text from PDF files using PyPDF2.
  • Splits extracted text into manageable chunks.

3. Embedding and Vectorization

  • Leverages Google Generative AI Embeddings for converting text into vectors.
  • Applies FAISS (Facebook AI Similarity Search) to create a vector store/index of text chunks.

4. Conversational Chain

  • Implements a conversational chain for question-answering using the Gemini Generative AI model.
  • Configures the prompt template for providing context and framing questions.

5. Workflow

  • Users upload PDF files and ask questions through the interface.
  • Text is extracted from PDFs, split into chunks, and converted into vectors.
  • The conversational chain processes user input, searches for similar text chunks, and generates responses.

Code Structure

  • main() function: Sets up the Streamlit interface and handles user input.
  • get_pdf_text(pdf_docs) function: Extracts text from PDF files.
  • get_text_chunks(text) function: Splits text into manageable chunks.
  • get_vector_store(text_chunks) function: Creates a vector store/index from text chunks.
  • get_conversational_chain() function: Configures the conversational chain for question-answering.
  • user_input(user_question) function: Processes user input and generates responses.
  • Environment variables: Utilizes the dotenv library to securely load the Google API key.

Usage

  1. Upload PDFs: Use the sidebar to upload one or more PDF files.
  2. Ask a Question: Enter your question in the provided text input.
  3. Submit & Process: Click the button to initiate the processing of PDFs and question-answering.
  4. View Response: The system generates a response based on the input question and the content of the PDFs.

Dependencies

  • Streamlit
  • PyPDF2
  • LangChain
  • Google Generative AI
  • FAISS
  • Dotenv

Setup

  1. Install Dependencies: Ensure the required Python packages are installed.
  2. Set up Google API Key: Store the Google API key in a secure manner using the dotenv file.
  3. Run the Application: Execute the script to launch the Streamlit interface.

Feel free to explore and enhance the functionalities of this project based on your requirements.

PDFGemini - Unleash the Power of Conversational PDF Exploration! 💬✨