The distribution of the model is not the end—it's only the beginning. The real problems start from here. We have no control over the data in the real environment. Changes may occur and we must be ready to detect and update our model before it becomes obsolete. Monitoring is important to ensure the reliability, availability, and performance of our machine learning application. In this recipe, we will discuss some tools that we can use to keep track of changes that occur in the model.
Keeping track of changes into production
How to do it...
The following tools are available to monitor an Amazon SageMaker application:
- Amazon CloudWatch: This tool, which is available in AWS, monitors the resources and applications...
