How Booking.com Serves Deep Learning Model Predictions
Offered By: EuroPython Conference via YouTube
Course Description
Overview
Explore the intricacies of deploying deep learning models in production at Booking.com in this 31-minute EuroPython 2017 conference talk. Delve into the challenges and solutions for serving model predictions, covering topics such as training models in Docker containers, automated retraining processes, and deployment using Kubernetes. Learn about optimizing prediction serving for both latency and throughput in a containerized environment. Gain valuable insights into the lifecycle of a model, from initial training on a laptop to full-scale production deployment. Discover practical applications in image tagging and recommendation engines, and understand how to overcome bottlenecks in putting machine learning models into production. Perfect for data scientists and engineers looking to bridge the gap between model development and real-world implementation.
Syllabus
Intro
Agenda
Image Tagging
Recommendation Engine
Lifecycle of a model
Training a Model - on laptop
Deploying a Model
Performance
Optimizing for Latency
Optimizing for Throughput
Summary
Want to get in touch?
Taught by
EuroPython Conference
Related Courses
Developing a Tabular Data ModelMicrosoft via edX Data Science in Action - Building a Predictive Churn Model
SAP Learning Serverless Machine Learning with Tensorflow on Google Cloud Platform 日本語版
Google Cloud via Coursera Intro to TensorFlow em Português Brasileiro
Google Cloud via Coursera Serverless Machine Learning con TensorFlow en GCP
Google Cloud via Coursera