metadata

title: DocuChat
emoji: 📖
colorFrom: green
colorTo: pink
sdk: docker
pinned: true
duplicated_from: mckplus/DocuChat

DocuChat: The LangChain-Powered PDF Analysis 🚀

Here's how it works:

Upload a PDF: Select any PDF you want to analyze.
Enter Your OpenAI API Key: To leverage OpenAI's powerful language models, you'll need to provide your API key.
Ask a Question: Type your question into the text box and hit "Run."
Get the Answer: The system will analyze your PDF and respond with the most relevant answer.

Technical Overview

PDF Loading and Parsing

PDF Loader: Utilizes the PyPDFLoader from LangChain to read the uploaded PDF file.
Document Splitter: The CharacterTextSplitter divides the PDF into manageable chunks for further processing.

Embeddings and Vector Storage

Embeddings: The OpenAIEmbeddings class is used to transform the text into a format that can be understood by the models.
Vector Storage: Chroma is used to store the vectors of the documents for quick retrieval.

Question Answering

RetrievalQA: This class, from LangChain's chains, handles the core of the question-answering process.
OpenAI as Language Model: The OpenAI class is utilized as the core language model for the retrieval.

Interactive UI

Panel Library: The entire interface is built using the Panel library, making it highly interactive and user-friendly.
Responsive Design: The UI adapts to various screen sizes, providing a smooth experience across devices.

Built by McKenzie

This project is created and maintained by McKenzie. Contributions and feedback are always welcome! & credit to sophiamyang for the inspo & intro to panel 🙏

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference