Explore the evolution of machine learning training and serving infrastructure at Pinterest Ads over a 5+ year period in this 26-minute conference talk from MLOps World. Discover how the company progressed from logistic regression-based models to large transformer-based models served using GPU technology. Learn about the challenges faced and lessons learned during this transformative journey that revolutionized Pinterest Ads' capabilities. Gain insights from Senior Machine Learning Engineer Aayush Mudgal as he shares the triumphs and tribulations encountered in this critical business undertaking.
Evolution of ML Training and Serving Infrastructure at Pinterest Ads
MLOps World: Machine Learning in Production via YouTube
Overview
Syllabus
Evolution of ML Training and Serving Infrastructure @ Pinterest Ads
Taught by
MLOps World: Machine Learning in Production
Related Courses
-
Ads Ranking Evolution at Pinterest - From Logistic Regression to Deep Learning
-
MLOps Platforms: Amazon SageMaker and Azure ML
-
AI Inference Workloads - Solving MLOps Challenges in Production
-
Production ML Serving and Monitoring with Kubernetes
-
Git-Based CI-CD for ML
-
Shopping Recommendations at Pinterest
