AiGptCode / AskyourDocuments

Welcome to the Document QA system! This repository contains the code for a system that allows you to ask questions about your documents and get answers based on their contents. It supports a wide range of document formats, including PDF, Word, Excel, PowerPoint, text files, and even images!

💻 Ask your Documents 🤖

👋 Welcome to the Document QA system! This repository contains the code for a system that allows you to ask questions about your documents and get answers based on their contents. It supports a wide range of document formats, including PDF, Word, Excel, PowerPoint, text files, and even images!

🚀 Features

• 💻 Supports a variety of document formats, including PDF, Word, Excel, PowerPoint, text files, and images
• 🤖 Uses the Hugging Face Transformers library to create embeddings for document chunks
• 🔍 Uses the FAISS library to create an index for those embeddings, allowing for efficient similarity search
• 💬 Allows users to ask questions about their documents and get answers based on the contents of those documents
• ⚡️ Uses multiprocessing to parallelize the creation of the index for improved performance

📋 Requirements

• Python 3.6 or higher
• The following Python packages:
  • transformers
  • langchain
  • fitz
  • Pillow
  • textract
  • pandas
  • python-pptx
  • concurrent-futures
  • opencv-python (for image support)

🔧 Usage

1. Clone this repository to your local machine:

git clone https://github.com/AiGptCode/AskyourDocuments.git

2. Install the required Python packages:

pip install transformers langchain fitz pillow textract pandas python-pptx opencv-python concurrent-futures

3. Set your Hugging Face API key as an environment variable:

export HUGGINGFACE_API_TOKEN=your-api-key

4. Run the main.py script and enter the path to the directory containing your documents:

python AskyourDocuments.py

5. Ask a question about your documents and get an answer based on the contents of those documents.

Note: If you want to include images in your search, make sure they are in a supported format (e.g., JPEG, PNG) and are located in the same directory as your other documents.

🤝 Contributing

If you would like to contribute to this project, please follow these steps:

1. Fork this repository to your own GitHub account.
2. Create a new branch for your changes:

git checkout -b my-feature-branch

3. Make your changes and commit them:

git commit -am 'Add some feature'

4. Push your changes to your fork:

git push origin my-feature-branch

5. Open a pull request against the original repository.

📄 License

This project is licensed under the MIT License.

🎉 Acknowledgments

• The Hugging Face Transformers library for providing pre-trained models and tokenizers
• The FAISS library for providing efficient similarity search and clustering of dense vectors
• The langchain library for providing utilities for creating and working with language models
• The fitz library for providing utilities for working with PDF files
• The Pillow library for providing utilities for working with image files
• The textract library for providing utilities for extracting text from various file formats
• The pandas library for providing utilities for working with tabular data in Python
• The python-pptx library for providing utilities for working with PowerPoint files
• The concurrent-futures library for providing a high-level interface for asynchronously executing callables
• The opencv-python library for providing utilities for working with image and video data (for image support)