Monitoring

Monitoring is one of the most important aspects of any machine learning workflow. Kaptain offers a range of metric dashboards that allow Kaptain users (for example data scientists) to see the current status of their machine learning workloads, their deployed models, and how many resources they are consuming. With the dashboards, users can observe the state of their ML workloads and models without having to piece together information from various sources.

There are two types of dashboards: the Kaptain dashboards are accessible via the Kubeflow UI, and the Grafana dashboards are accessible via the Grafana UI.