GitOps for Machine Learning Pipelines - Rishit Dagli, University of Toronto & Shivay Lamba
Description:
Explore how GitOps principles can be applied to Machine Learning pipelines in this informative conference talk. Learn about the core processes of an ML lifecycle and discover Kubeflow, an open-source project that simplifies Machine Learning on Kubernetes. Gain insights into leveraging GitOps for various ML processes, including training experiments, hyperparameter tuning, end-to-end pipelines, and training jobs. Understand how to deploy multiple Kubeflow components using Argo CD and implement GitOps practices for ML tasks. Conclude with a demonstration of deploying an end-to-end Kubeflow ML pipeline using GitOps techniques, equipping you with practical knowledge to enhance your ML workflows.