manuthvann216 / chatwithpdf_openai

This project allows you to upload a PDF document and ask questions about its content. It uses langchain, openai api model and Facebook Ai Similarity Search(FAISS) library to process the text in the PDF and provide answers to questions pertaining the document.

chatWithPDF

This project allows you to upload a PDF document and ask questions about its content. It uses langchain, openapi ai model and Facebook Ai Similarity Search(FAISS) library to process the text in the PDF and provide answers to questions pertaining the document.

Project Screen

Cost

Installation

1. Clone the repository:

   git clone https://github.com/john-thuo1/chatWithPDF
   cd into your directory/ open with vscode

2. Create a Virtual Environment:

   python -m venv env

3. Install the required dependencies:

   pip install -r requirements.txt

4. Create OpenAI API Key and add it to your .env file: openai

5. Run the application:

   streamlit run App.py

Next Steps

1. Add support for multiple file formats
2. Implement Document Indexing techniques by use of libraries such as Elasticsearch or Apache Solr
3. Enhance question answering capabilities: Explore advanced question answering techniques, such as using transformer models like BERT or GPT, to improve the accuracy and comprehension of the system.
4. Use a model that supports multiple languages, most notably some BERT models do support this.