While working on your ML experiments, you are setting up compute resources which you want to share with two of your colleagues in your workspace.
For sharing computer resources, which of the following statements are true?
Click on the arrows to vote for the correct answer
A. B. C. D.Correct Answer: D.
Option A is incorrect because as a development environment, compute instance cannot be shared with other users in the workspace.
Option B is incorrect because compute resources, except for development compute instances, can be shared among multiple users.
Option C is incorrect because the compute instance cannot be shared, while the compute cluster can.
Option D is CORRECT because as a development environment, compute instances cannot be shared with other users in the workspace, while training clusters can be used in shared mode to run jobs launched by several users.
References:
The correct answer is A. You can share your compute instance for development; you can share your compute cluster for training.
When working on ML experiments, compute resources such as virtual machines and clusters are often required to process large amounts of data and train complex models. Sharing these resources with colleagues in a workspace can help to increase collaboration and productivity.
In Azure Machine Learning, there are two types of compute resources that can be used: compute instances and compute clusters.
A compute instance is a single virtual machine that is used for development and testing. It provides a Jupyter Notebook interface that allows data scientists to write and execute code in a familiar environment. Compute instances can be shared with colleagues by providing them with access to the workspace, or by creating a custom role that allows them to access the instance.
A compute cluster, on the other hand, is a group of virtual machines that are used for distributed training and hyperparameter tuning. Compute clusters can be scaled up or down depending on the size of the workload, and they are typically used for more computationally-intensive tasks. Compute clusters can also be shared with colleagues by providing them with access to the workspace, or by assigning them to a shared compute target.
Based on the above, it is clear that statement A is correct. Compute instances can be shared for development, and compute clusters can be shared for training. Statements B and C are incorrect because they suggest that sharing compute resources is not possible, which is not true. Statement D is incorrect because it suggests that compute instances cannot be shared, which is not true.