Accelerate Deep Learning Workloads with Amazon SageMaker

4

Managing Deep Learning Datasets

Deep learning models usually require a considerable amount of training data to learn useful patterns. In many real-life applications, new data is continuously collected, processed, and added to the training dataset, so your models can be periodically retrained so that they can adjust to changing real-world conditions. In this chapter, we will look into SageMaker capabilities and other AWS services to help you manage your training data.

SageMaker provides a wide integration capability where you can use AWS general-purpose data storage services such as Amazon S3, Amazon EFS, and Amazon FSx for Lustre. Additionally, SageMaker has purpose-built storage for machine learning (ML) called SageMaker Feature Store. We ...