Using Kubeflow to Run AI/MLOps Workloads
In the previous chapter, we looked at several logging, monitoring, and alerting options to gain comprehensive visibility into our container infrastructure and workloads. Regarding tools for setting up a monitoring and alerting stack, we looked at Prometheus, Grafana, and Alert Manager. We also looked at how to use the EFK toolset to set up a centralized, cluster-level logging stack that can handle large volumes of log data. Finally, we discussed the key indicators to keep a close eye on so that you can effectively manage your infrastructure and applications.
In this chapter, we will go through the steps for creating a machine learning (ML) pipeline that will build and deploy a sample ML model using the Kubeflow MLOps platform. ML is an AI subfield. The purpose of ML is to teach computers to learn from the data you provide. Instead of describing the action, the machine will take your code and provide an algorithm that adjusts, depending on...
