README.md

---
title: Paper Reader Assistant
emoji: 📚
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: "4.44.0"
app_file: app.py
pinned: false
---

# 📖 Paper Reader Assistant (Gradio on Hugging Face Spaces)

This Space helps you **read and analyze academic papers** with AI assistance.  
You can upload PDF files, and the app will automatically extract text, generate summaries, identify keywords, and enable question answering using RAG (Retrieval-Augmented Generation).

---

## 🚀 Features

- 🧾 **PDF Upload** → Extracts text using PyMuPDF  
- ✍️ **Summarization** → Uses BART for map-reduce summarization  
- 🔑 **Keyword Extraction** → Powered by KeyBERT  
- ❓ **Q&A (RAG)** → Combines FAISS + SentenceTransformers + Flan-T5  
- 🧠 **All-in-one academic reading assistant**  

---

## ⚙️ Quick Start

1. Create a new **Space** on Hugging Face with SDK = **Gradio**  
2. Upload three files:  
   - `app.py`  
   - `requirements.txt`  
   - `README.md` (this file)  
3. Save changes — Spaces will automatically build and launch the app  
4. Once it’s live, open the Space URL to start using it  

---

## 💡 Notes

- Works best on **English scientific PDFs** (text-based).  
- For **scanned PDFs**, run OCR first (not yet included in this version).  
- Models run on **CPU by default** — enable GPU for faster processing.  
- You can further customize UI theme, font color, or add dark mode by editing `app.py`.

---

## 🧰 Tech Stack

| Component | Description |
|------------|-------------|
| **Gradio** | Front-end interface |
| **PyMuPDF** | PDF text extraction |
| **Transformers (BART)** | Summarization |
| **KeyBERT** | Keyword extraction |
| **Sentence-Transformers + FAISS** | Vector retrieval for Q&A |
| **Flan-T5** | Natural language generation |

---

**Author:** WENIor  
**License:** MIT  
**Version:** 1.0.0