0

Reading through Azure documentation on Cognitive Service LUIS and Text Analytics, and it only mentioned the on-prem container offers the "same capabilities", but not mentioning if it is the same model, in terms of quality and deterministic behaviors.

Any model quality difference between Azure Cognitive Service container vs Cloud that someone could offer any insights?

Lewis Liu
  • 63
  • 1
  • 5

1 Answers1

0

As Azure cognitive service official indicated :

Cognitive Services containers are an alternative to the Cognitive Services cloud.Cognitive Services containers are an alternative to the Cognitive Services cloud. Containers offer the same capabilities as the corresponding cloud services. Customers can deploy the containers on-premises or in Azure. The core AI technology, pricing tiers, API keys, and API signature are the same between the container and the corresponding cloud services.

So there will be no differences between between Azure Cognitive Service container and Cloud. And Azure Cognitive Service container provides customers flexibility in versioning and updating of models deployed in their solutions

If your data can not be used on Azure due to some privacy or regulatory restrictions or you are using massive data loads, which would either be too expensive or time-consuming to bring to the cloud or High throughput / low latency is required ,using Azure Cognitive Service container will be a great practice.

Stanley Gong
  • 11,522
  • 1
  • 8
  • 16
  • I saw the same message. It says same "capabilities", but not sure about if it is the same quality model, or a different model for on-prem use. – Lewis Liu Nov 19 '19 at 06:36