guardrails demo

LLM Security Project with Llama Guard

This repository provides a quick and easy way to run the Llama Guard application on your local machine and explore LLM security.

What is Llama Guard?

Llama Guard is a defensive framework designed to detect and mitigate potential security risks associated with Large Language Models (LLMs). It helps developers and researchers build safer and more reliable LLM applications.

What's included?

Nemo Guardrail Implementation: The llama-guard folder contains a NeMo Guardrail implementation, offering flexibility and customization for your specific needs. Streamlit Applications: Two Streamlit applications are provided for convenient testing: llama-guard-only.py: Test input prompts and responses directly with Llama Guard. llama_2_with_llama-guard.py: Run Llama Guard with the pre-trained Llama 2 13b model for real-world testing.

How to run the application:

• Clone this repository
• Install dependencies: pip install -r requirements.txt
• Run the desired application:
  1. Test Llama Guard: streamlit run llama-guard-only.py
  2. Test with Llama 2 13b: streamlit run llama_2_with_llama-guard.py

Learn More:

Blog post: Deepen your understanding of Llama Guard and LLM security with this informative blog post: https://balavenkatesh.medium.com/securing-tomorrows-ai-world-today-llama-guard-defensive-strategies-for-llm-application-c29a87ba607f