Pluralsightintermediateen

Production Machine Learning Systems

Learn to build production-ready machine learning systems. Covers static/dynamic inference, continuous training, and batch/online processing. Master ML deployment!

2h 43m
Certificate
Google Cloud
Google CloudInstructor

Course Content

Production Machine Learning Systems is an intermediate-level Pluralsight course taught by Google Cloud. It runs 2h 43m, is taught in English, and includes a certificate of completion. It is a paid course, first published December 2022.

Master the deployment of production machine learning systems with this advanced course, the second in the Advanced Machine Learning on Google Cloud series. You'll gain practical skills in building and managing ML pipelines for a variety of use cases.

Key topics include:

  • Production ML System Flavors: Explore and implement static, dynamic, and continuous training strategies to adapt to evolving data and model requirements.
  • Inference Strategies: Learn to deploy models for static and dynamic inference, optimizing for latency and throughput.
  • Batch vs. Online Processing: Understand the differences and choose the right processing approach for your specific ML application.
  • TensorFlow Abstraction Levels: Deep dive into TensorFlow's abstraction layers to build scalable and maintainable models.
  • Distributed Training: Discover various options for distributed training on Google Cloud, maximizing compute resources and accelerating model development.
  • Custom Estimators: Learn how to write distributed training models using custom estimators for maximum flexibility and control.

This course provides the knowledge and skills necessary to deploy robust and scalable machine learning solutions on Google Cloud.

Next Steps: After completing this course, advance your expertise with the Image Understanding with TensorFlow on Google Cloud course.

$ whoami — the devs behind tutorialsearch.io

Want AI in your product? Hire the team behind TutorialSearch.io

LLM features, RAG, agents and automation wired into a real product — built by the devs who shipped this site.