Master Secure LLMs with NVIDIA Guardrails Pro
Focused View
56:49
01 - Leverage guardrails to build secure LLMs.mp4
01:08
02 - What you should know.mp4
01:02
03 - Exercise files and setting up your environment.mp4
01:43
01 - Introduction to guardrails for LLMs.mp4
01:55
02 - Behind the scenes How Guardrails enforces LLM safety.mp4
04:35
03 - Understanding the role of embeddings in safeguarding LLMs.mp4
03:10
01 - Defining conversational boundaries.mp4
02:52
02 - Advanced Colang flows.mp4
04:34
03 - Guardrails in action.mp4
02:42
01 - Enhancing LLMs with custom actions.mp4
03:40
02 - Building a custom inquiry action.mp4
03:20
01 - Retrieval-augmented generation (RAG) with actions.mp4
05:09
02 - Securing LLMs against sensitive topics.mp4
04:01
03 - Debugging and optimizing Guardrails.mp4
04:18
01 - Case studies Guardrails in action.mp4
02:49
02 - Best practices for implementing guardrails.mp4
05:23
03 - Future of guardrails and LLM safety.mp4
02:46
01 - Next steps in building LLM applications.mp4
01:42
More details
Course Overview
Explore how NVIDIA NeMo Guardrails safeguards LLMs against misuse while enhancing conversational standards and AI trust. Learn to implement ethical AI deployment, craft dynamic interactions, and leverage RAG for accurate responses.
What You'll Learn
- Enforce LLM safety using NVIDIA Guardrails
- Design conversational boundaries with Colang
- Augment LLMs with custom actions and RAG
Who This Is For
- AI practitioners prioritizing ethical deployment
- Developers building secure LLM applications
- Technology advocates focused on AI safety
Key Benefits
- Prevent LLM misuse and sensitive topic exposure
- Elevate response quality with retrieval-augmented generation
- Implement industry best practices through real-world case studies
Curriculum Highlights
- Foundations of Guardrails for Language Models
- Crafting Conversational Guardrails
- Integrating Custom Actions
Focused display
Category
- language english
- Training sessions 18
- duration 56:49
- English subtitles has
- Release Date 2025/06/07
