Deploying a Model for Inference at Production (NVIDIA) Course Overview

Course Prerequisites

Target Audience for Deploying a Model for Inference at Production (NVIDIA)

Deploying a Model for Inference at Production Scale (NVIDIA) course is designed for professionals looking to efficiently scale machine learning models using NVIDIA Triton Inference Server and Prometheus, ensuring robust performance at production levels.

• Machine Learning Engineers
• Data Scientists
• AI/Machine Learning Researchers
• Software Engineers specializing in AI/ML
• DevOps Engineers working with MLOps
• AI/ML Project Managers
• Technology Consultants in AI/ML
• Technical Leads in AI/ML Projects
• Machine Learning Infrastructure Engineers
• IT Architects focused on AI solutions

Learning Objectives - What you will Learn in this Deploying a Model for Inference at Production (NVIDIA)?

Introduction

The "Deploying a Model for Inference at Production Scale" (NVIDIA) course equips learners with the skills to deploy and scale machine learning models effectively using NVIDIA Triton Inference Server and Prometheus, focusing on practical applications and metrics analysis.

Learning Objectives and Outcomes

• Deploy neural networks from various frameworks onto a live Triton Server.
• Measure GPU usage and other essential metrics with Prometheus.
• Send asynchronous requests to maximize throughput.
• Implement PyTorch-based Convolutional Neural Networks (CNNs).
• Apply data augmentation techniques for model improvement.
• Utilize transfer learning for enhanced model performance.
• Integrate natural language processing models.
• Start with setting up and running simple models using PyTorch, TensorFlow, TensorRT, and HuggingFace.
• Develop advanced inference strategies.
• Track and analyze performance metrics for optimized deployment.

Upon completion, learners will be capable of deploying their own machine learning models on a GPU server efficiently.