MLOps: DevOps for AI with KubeFlow Pipelines
summary
Overview of MLOps practices (applying DevOps principles to the ML lifecycle) and a demonstration of Kubeflow Pipelines. It covers how Kubeflow Pipelines on Kubernetes enables scalable, repeatable, cloud-oriented ML workflows to improve agility, collaboration between data scientists and ML engineers, and consistency/reliability across development and deployment.
AI-written summary · about
chapters (8)
- 0:00MLOps intro
- 3:08Tool selection
- 4:54Model dev myths
- 6:59CI/CD deployment
- 7:41Kubeflow Pipelines features
- 8:45Kubeflow demo
- 11:34MLOps use cases
- 12:58Getting started with Kubeflow
related talks
05mCI/CD Pipelines and Machine Learning
25mWhen Kubernetes Meets MLflow: Orchestrating MLOps Processes the Cloud Native Way
22mBuilding a MLOps Platform with Kubeflow on GKE
20mML Ops design patterns with Kubeflow Pipelines
31mKubernetes and MLOps for Scalable and Reproducible Generative AI
36mIs the MLOps tooling the same as the DevOps tooling?
26mKubernetes and MLOps for Scalable and Reproducible Generative AI
30mEnd-to-End MLOps with MLflow and Kubeflow