Part 3 – Deployment and Maintenance

Many complex challenges often arise during deployment of a deep learning project. In many cases, the deployment settings are different from the development settings and the discrepancy can introduce various restrictions. In this part, we introduce common issues that engineers often struggle with and share effective solutions for each challenge. In the final chapter, we describe the last phase of a deep learning project, which consists of evaluating the project and discussing potential improvements for future projects.

This part comprises the following chapters:

• Chapter 8, Simplifying Deep Learning Model Deployment
• Chapter 9, Scaling a Deep Learning Pipeline
• Chapter 10, Improving Inference Efficiency
• Chapter 11, Deep Learning on Mobile Devices
• Chapter 12, Monitoring Deep Learning Endpoints in Production
• Chapter 13, Reviewing the Completed Deep Learning Project