what is the difference between azure-ail-ml and azureml-sdk package?

Question

I'm was wondering what is the purpose of azure-ai-ml package as azureml-sdk already provides the api to create everything machine learning related. For context, I have experience building ML solutions using azureml.core and it's just that doing the same thing using azure.ai.ml feels very different.

Lastly, for giving the DP-100 certification exam which one should I learn? From what I gather from the DP-100 learning path, it looks like azure.ai.ml is used

TIA

Answer 1

@Das Gupta, Abhijeet Thanks for the question, The azure-ai-ml package is part of the Azure Machine Learning Python SDK v2, while azureml-sdk is part of the Azure Machine Learning Python SDK v1. The key difference between these two versions is the consistency of features and terminology across the interfaces.

In order to create this consistency, the syntax of commands differs, in some cases significantly, from the first versions (v1). There are no differences in functionality between SDK v2 and CLI v2. The command line based CLI may be more convenient in CI/CD MLOps type of scenarios, while the SDK may be more convenient for development.

As for your DP-100 certification exam, it seems that the exam will be updated on October 18, 2023. If you plan to take the exam after that date, you should review the skills measured as of October 18, 2023. The study guide linked in the “Tip” box on the official exam page provides details about the skills measured and upcoming changes.

https://learn.microsoft.com/en-us/credentials/certifications/exams/dp-100/