🐍 AI Python Tutor (Local RAG · LLaMA-2)

An interactive Python coding tutor powered by local RAG (Retrieval-Augmented Generation).
Upload your own tutorials (PDF, TXT, or Markdown), and ask coding questions.
The app retrieves relevant snippets, feeds them to a language model, and returns clear, runnable Python code with inline comments.

---

✨ Features

• 📂 Reads and indexes local .pdf, .txt, and .md files.
• 🔎 Fast and efficient semantic search with FAISS.
• 🧠 Uses HuggingFace embeddings (all-mpnet-base-v2) for vectorization.
• 🤖 LLM-powered tutor (supports LLaMA-2, Ollama, or OpenAI/Gemini).
• 🐍 Always returns complete runnable Python code with comments.
• 📚 Expandable source panel to see where answers come from.
• 🎨 Clean Streamlit UI with background customization.

---

🚀 Getting Started

1️⃣ Clone the repo

git clone https://github.com/yourusername/ai-python-tutor.git
cd ai-python-tutor
python -m venv venv
source venv/bin/activate   # Linux/Mac
venv\Scripts\activate      # Windows
streamlit run app.py

⚙️ Configuration

Dataset folder → set in the sidebar.

Temperature → controls creativity (0.0 = precise, 1.0 = more creative).

Build/Rebuild Knowledge Base → (re)creates FAISS vector store.