Overview
Explore the critical role of Prometheus in the MLOps lifecycle through this informative conference talk. Gain a comprehensive understanding of monitoring ML deployments, including edge cases in production, data drift, concept drift, and model metrics. Discover how to effectively monitor and manage ML models in production environments using Prometheus. Learn techniques for enhancing existing deployments with powerful monitoring capabilities and explore practical demonstrations of integrating Prometheus with Flyte, Seldon Core, and FastAPI ML deployments. Acquire valuable insights into determining when to retrain models, collect additional data, or adjust data collection strategies based on monitoring results.
Syllabus
Prometheus in the MLOps Lifecycle - Rishit Dagli & Shivay Lamba
Taught by
CNCF [Cloud Native Computing Foundation]