Udostępnij za pośrednictwem


NEW REFERENCE ARCHITECTURE: Batch scoring of Python models on Azure

Our fifth AI Reference Architecture (on the Azure Architecture Center) comes from AzureCAT Said Bleik.

This reference architecture shows how to build a scalable solution for batch scoring many models on a schedule in parallel using Azure Batch AI.

The solution monitors the operation of a large number of devices in an IoT setting where each device sends sensor readings continuously.

This Reference Architecture includes the following information:

  1. Architecture - Explaining the different elements of the architectural diagram.
  2. Performance considerations - A deeper look at parallelizing across virtual machines, cores, and file servers.
  3. Management considerations - Notes on monitoring Batch AI jobs and logging in Batch AI.
  4. Cost considerations - Examining compute resources, Batch AI cluster sizes, and other scaling factors.
  5. Deploy the solution - Our GitHub repo features more details, the prerequisites, setup steps, deployment instructions, along with the scripts and commands.

Head over to the Azure Architecture Center to learn more about this reference architecture, Batch scoring of Python models on Azure.

 

See Also

Additional related AI Reference Architectures:

Find all our Reference Architectures here.

 

AzureCAT Guidance

"Hands-on solutions, with our heads in the Cloud!"