Exam DP-100 topic 1 question 39 discussion

Should be True. Because we can use databricks or vm as attached compute for training purposes

Azure Machine Learning designer to train models. its the key so answer is compute cluster.

Attached compute is already supported by Azure Machine Learning Designer

There are now 5 tabs in compute, and "Attached computes" is one of them. When you configure & submit a pipeline job, you can choose attached compute.

A-Yes is answer

We can use attached compute. Just tried it

Correct Answer: B See the blue box in the link below which says 'Attached compute is not supported, use compute instances or clusters instead.' https://learn.microsoft.com/en-us/azure/machine-learning/tutorial-designer-automobile-price-train-score

It is not possible to determine whether the requirements will be satisfied based on the information provided. The recommendation to choose Attached compute is a specific solution, but the requirements are not stated. It is important to have a clear understanding of the requirements and constraints before making a decision on which compute type to use in Azure Machine Learning Designer.

ChatGPT says A

https://learn.microsoft.com/en-us/azure/machine-learning/concept-designer#compute

As is shown in the links below, Designer only supports Azure Machine Learning Compute (AML Compute cluster or AML compute instance) for training. 1. https://learn.microsoft.com/en-us/azure/machine-learning/concept-designer#compute 2. https://learn.microsoft.com/en-us/azure/machine-learning/concept-compute-target#training-compute-targets

Cannot use Attached Compute Cluster in designer: https://docs.microsoft.com/en-us/azure/machine-learning/concept-compute-target Same context but slightly different question on exam 07/11/2022

Answer is A

No is the correct answer. https://learn.microsoft.com/en-us/azure/machine-learning/concept-compute-target ML Designer support 3 target: 1.Azure Machine Learning compute cluste 2.Azure Machine Learning compute instance 3.Azure Machine Learning Kubernetes

i just tried in AML designer, can choose attached compute now

can attach Kubernetes service for training and inferencing (preview) Azure Machine Learning provides you with the following options to attach your own Kubernetes clusters for training and inferencing: Azure Kubernetes Service. Azure Kubernetes Service provides a managed cluster in Azure. Azure Arc Kubernetes. Use Azure Arc-enabled Kubernetes clusters if your cluster is hosted outside of Azure. Those 2 also listed in the https://docs.microsoft.com/en-us/azure/machine-learning/concept-compute-target

The correct answer should have been A