🤖 SLM AI Assistant (GPU Accelerated)

A locally hosted Small Language Model (SLM) AI Assistant built using Python, Flask, and open-source LLMs, with optional RAG (Retrieval-Augmented Generation).

---

🚀 Features

• 🔹 Local AI assistant (no cloud APIs)
• 🔹 GPU-accelerated inference (NVIDIA CUDA)
• 🔹 Supports lightweight SLMs (Qwen-0.5B, Phi-2)
• 🔹 RAG using PDF documents
• 🔹 FAISS-based vector search
• 🔹 Flask REST API backend
• 🔹 Chat-style web UI
• 🔹 Optimized for low latency on consumer GPUs

---

🧠 Tech Stack

• Python
• Flask
• HuggingFace Transformers
• PyTorch + CUDA
• FAISS
• Sentence Transformers
• NVIDIA GPU Acceleration

---

⚙️ Setup Instructions

1️⃣ Clone the repository

git clone https://github.com/your-username/SLM-AI-Assistant.git
cd SLM-AI-Assistant

2️⃣ Install dependencies

python setup.py

3️⃣ Run the app

python app.py

Open in browser:

http://localhost:5000