This 14-minute tutorial from Nvidia demonstrates how to customize AI agents for precise function calling using NVIDIA NeMo microservices. Discover an end-to-end approach for maintaining efficiency and accuracy as available tools scale from tens to hundreds. The video covers essential prerequisites, data preparation techniques for fine-tuning and evaluation, model customization through LoRA fine-tuning, accuracy evaluation methods for customized models, and implementing guardrails to safeguard LLM behavior. Perfect for developers working with generative AI, LLM fine-tuning, custom generative AI solutions, AI agents, and agentic AI applications. The tutorial includes links to additional resources including a technical blog and downloadable NeMo microservices to support implementation.
Customizing AI Agents for Tool Calling with NVIDIA NeMo Microservices
Overview
Syllabus
00:00 – Introduction
00:54 – Prerequisites
02:57 – Preparing Data for Fine-Tuning and Evaluation
05:45 – Customizing the Model with LoRA Fine-Tuning
08:52 – Evaluating the Accuracy of the Customized Model
11:22 – Adding Guardrails to Safeguard Your LLM Behavior
Taught by
NVIDIA Developer
