RAG LLM Project

This project implements a simple Retrieval-Augmented Generation (RAG) system using:

• Hugging Face Transformers (LLM - Mistral)
• SerpAPI (Google Search for retrieval)

---

1. Clone the Repository

git clone https://github.com/cheikne/rag-llm.git cd rag-llm

Create Virtual Environment

python -m venv venv source venv/bin/activate

Install Dependencies

pip install --upgrade pip pip install torch torchvision torchaudio pip install -r requirements.txt

Set Environment Variables

SerpAPI Key (Google Search)

export SERPAPI_API_KEY="your_serpapi_key"

Hugging Face Token (for private models like Mistral)

export HF_TOKEN="your_hf_token"

If you wanna make Variables Permanent (Mac/Linux)

echo 'export SERPAPI_API_KEY="your_serpapi_key"' >> ~/.zshrc echo 'export HF_TOKEN="your_hf_token"' >> ~/.zshrc source ~/.zshrc

Run the Project

python main.py