Article complete

NVIDIA is offering a four-hour, self-paced course on MLOps

NVIDIA Triton helps data scientists and system administrators turn models into a web server for model prediction using the same training machine

NVIDIA is offering a four-hour, self-paced course on MLOps

Key takeaways

  • NVIDIA is offering a self-paced course focused on MLOps and deploying models using Triton Inference Server.
  • The course includes practical experience with deploying neural networks from various frameworks onto Triton Server.
  • Participants will learn to measure GPU usage and send asynchronous requests to enhance throughput.
  • Completion of the course enables developers to deploy their models on NVIDIA Triton Server effectively.
  • Triton Inference Server allows for model prediction using training machines, optimising performance with multiple GPUs.

Takeaways drafted with AI from the article text.

The course's learning objectives include: 

  • Deploying neural networks from a variety of frameworks onto a live NVIDIA TritonServer.
  • Measuring GPU usage and other metrics with Prometheus.
  • Sending asynchronous requests to maximise throughput.

Upon completion, developers will deploy their models on an NVIDIA Triton Server.

What is Triton Server?

NVIDIA Triton was created with Machine Learning Operations (MLOps). This is a relatively new field evolved from Developer Operations, or DevOps, to focus on scaling and maintaining ML models in a production environment. 

NVIDIA Triton is equipped with features such as model versioning for easy rollbacks. Triton is also compatible with Prometheus to track and manage server metrics such as latency and request count.

Published Updated

Spotted an error? Tell our editors. Read our editorial standards and corrections.