Amazon SageMaker launches Managed Spot Training for saving up to 90% in machine learning training costs
Amazon SageMaker now supports a new fully managed option called Managed Spot Training for training machine learning models using Amazon EC2 Spot instances. Spot Instances let you take advantage of unused compute capacity in the AWS cloud, and as a result, you can optimize the cost of training machine learning models by up to 90% compared to on-demand instances.
Amazon SageMaker now supports a new fully managed option called Managed Spot Training for training machine learning models using Amazon EC2 Spot instances. Spot Instances let you take advantage of unused compute capacity in the AWS cloud, and as a result, you can optimize the cost of training machine learning models by up to 90% compared to on-demand instances.
Amazon SageMaker manages Spot instances on your behalf, so you don’t have to poll continuously for capacity. There is no need to build additional tooling as Amazon SageMaker enables your training jobs to run reliably as and when Spot capacity becomes available. Managed Spot Training can be used when training models built using the popular ML frameworks in SageMaker, SageMaker built-in algorithms, and custom built models. You can also use Managed Spot Training Automatic Model Tuning to tune your machine learning models.
Managed Spot Training is supported for all instance types available in Amazon SageMaker and in all AWS regions where Amazon SageMaker is available. You can refer to the AWS documentation for details and also learn more in the blog post.