Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/sujal-maheshwari2004/talk-with-your-docs
A LLM powered chatbot with RAG and BS4 implementation for increased usage range made easier to use with a sleek StreamLit UI
https://github.com/sujal-maheshwari2004/talk-with-your-docs
embeddings llm openai-api python rag streamlit
Last synced: 13 days ago
JSON representation
A LLM powered chatbot with RAG and BS4 implementation for increased usage range made easier to use with a sleek StreamLit UI
- Host: GitHub
- URL: https://github.com/sujal-maheshwari2004/talk-with-your-docs
- Owner: sujal-maheshwari2004
- Created: 2024-07-16T12:05:03.000Z (7 months ago)
- Default Branch: main
- Last Pushed: 2024-10-31T04:31:48.000Z (3 months ago)
- Last Synced: 2024-12-02T00:16:18.634Z (2 months ago)
- Topics: embeddings, llm, openai-api, python, rag, streamlit
- Language: Python
- Homepage:
- Size: 17.6 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Chat with Docs
This is a Streamlit-based application that allows users to interact with PDF documents by querying content from uploaded PDFs or retrieving information from Wikipedia articles. Using embeddings and a question-answering model, users can ask questions about the document content for quick and contextual answers.
## Features
- **PDF Upload**: Upload multiple PDF files locally and ask questions based on their content.
- **Wikipedia Integration**: Search and download Wikipedia articles as PDFs, then ask questions about the article content.
- **Embeddings and Query System**: Uses embeddings to generate contextually relevant answers to user questions from the document content.
- **Downloadable Wikipedia PDFs**: Save Wikipedia article content as a PDF for offline use.
## Installation
### Prerequisites
- Python 3.8 or above
- [Streamlit](https://streamlit.io/)
- [PyPDF2](https://pypi.org/project/PyPDF2/)
- [FPDF](https://pypi.org/project/fpdf/)
- [BeautifulSoup](https://pypi.org/project/beautifulsoup4/)
- [Requests](https://pypi.org/project/requests/)
- [LangChain](https://pypi.org/project/langchain/)
- [FAISS](https://pypi.org/project/faiss-cpu/) (for local CPU-based usage)
- [dotenv](https://pypi.org/project/python-dotenv/)
- [Wikipedia-API](https://pypi.org/project/wikipedia-api/)
### Setting up the Project
1. **Clone the repository**:
```bash
git clone https://github.com/sujal-maheshwari2004/Talk-With-Your-Docs.git
cd chat-with-docs
```
2. **Install dependencies**:
```bash
pip install -r requirements.txt
```
3. **Set up OpenAI API Key**:
- Create a `.env` file in the root directory.
- Add your OpenAI API key:
```
OPENAI_API_KEY=your_openai_api_key
```
4. **Run the Application**:
```bash
streamlit run app.py
```
## Usage
1. **Select PDF Source**: Choose between uploading a local PDF file or searching for a Wikipedia article.
2. **Local PDF Upload**: Upload one or multiple PDF files, ask questions, and receive answers from the content.
3. **Wikipedia Article**: Enter a search query to retrieve a Wikipedia article, download it as a PDF, and ask questions about it.
## File Structure
- `app.py`: Main application script for the Streamlit interface.
- `wiki_utils.py`: Contains helper functions for searching and scraping Wikipedia pages.
- `pdf_utils.py`: Contains functions to save data to PDF and extract text from PDFs.
- `embeddings_utils.py`: Contains the function `create_embeddings_and_query` to generate embeddings for document content and answer user queries.
## Dependencies
Make sure to check `requirements.txt` for a full list of dependencies.
## Future Improvements
- Add support for more document formats (e.g., Word documents).
- Enhance question-answering capabilities by experimenting with different embeddings and model types.
- Add caching for previously queried content to improve performance.