Building Secure and Trustworthy LLMs Using NVIDIA Guardrails

Course details

Guardrails are essential components of large language models (LLMs) that can help to safeguard against misuse, define conversational standards, and enhance public trust in AI technologies. In this course, instructor Nayan Saxena explores the importance of ethical AI deployment to understand how NVIDIA NeMo Guardrails enforces LLM safety and integrity. Learn how to construct conversational guidelines using Colang, leverage advanced functionalities to craft dynamic LLM interactions, augment LLM capabilities with custom actions, and elevate response quality and contextual accuracy with retrieval-augmented generation (RAG). By witnessing guardrails in action and analyzing real-world case studies, you'll also acquire skills and best practices for implementing secure, user-centric AI systems. This course is ideal for AI practitioners, developers, and ethical technology advocates seeking to advance their knowledge in LLM safety, ethics, and application design for responsible AI.

Concepts

0. Introduction

01 - Leverage guardrails to build secure LLMs
02 - What you should know
03 - Exercise files and setting up your environment

1. Foundations of Guardrails for Language Models

04 - Introduction to guardrails for LLMs
05 - Behind the scenes - How Guardrails enforces LLM safety
06 - Understanding the role of embeddings in safeguarding LLMs

2. Crafting Conversational Guardrails

07 - Defining conversational boundaries
08 - Advanced Colang flows
09 - Guardrails in action

3. Integrating Custom Actions

10 - Enhancing LLMs with custom actions
11 - Building a custom inquiry action

4. Retrieval Augmented Generation (RAG) with Guardrails

12 - Retrieval-augmented generation (RAG) with actions
13 - Securing LLMs against sensitive topics
14 - Debugging and optimizing Guardrails

5. Real-World Applications and Best Practices

15 - Case studies - Guardrails in action
16 - Best practices for implementing guardrails
17 - Future of guardrails and LLM safety

Conclusion

18 - Next steps in building LLM applications