Google Professional Machine Learning Engineer Question 148
Single answerGoogle Cloud PlatformYou are a machine learning engineer tasked with building a custom model for image classification. You want to use Jupyter notebooks to prototype your solution and train the model. The solution requires access to GPUs for accelerated training. Which Jupyter backend should you choose on Google Cloud to meet your requirements?
- A
Vertex AI Workbench (User-Managed Notebooks)
- B
Vertex AI Workbench (Managed Notebooks) with GPU enabled
- C
Colab Notebooks
- D
Google Cloud Console Notebook Viewer
Show answer and explanation
Correct answer: B
Explanation
Vertex AI Workbench (Managed Notebooks) with GPU enabled is the best choice for this scenario. It provides a managed, scalable Jupyter environment with GPU support, making it easy to prototype and train machine learning models on Google Cloud. Other options either lack GPU support, require manual setup, or are not suitable for running notebooks.
- A. Incorrect.
Vertex AI Workbench (User-Managed Notebooks) can support custom environments and GPU acceleration, but it requires manual setup and management. While viable, it is not the most straightforward option for this requirement.
- B. Correct.
Vertex AI Workbench (Managed Notebooks) with GPU enabled is a fully managed Jupyter backend that provides pre-configured environments with GPU support. It is the most appropriate choice for this scenario as it minimizes setup time and provides seamless integration with Google Cloud services.
- C. Incorrect.
Colab Notebooks is a free offering with limited support for custom cloud resources and lacks native integration with Google Cloud for advanced configurations like GPU quotas. While useful for prototyping, it is not ideal for this use case.
- D. Incorrect.
Google Cloud Console Notebook Viewer is a tool for viewing notebook files but does not support running notebooks or leveraging GPUs for model training.