Google Professional Machine Learning Engineer Question 145
Single answerGoogle Cloud PlatformYou are a machine learning engineer working on a project that involves training a large deep learning model. The model requires GPU acceleration for efficient training. You want to use a managed Jupyter Notebook environment on Google Cloud that supports seamless integration with other Vertex AI services. Which Jupyter backend should you choose?
- A
Vertex AI Workbench (Managed Notebooks)
- B
Google Colab
- C
Custom Compute Engine VM with JupyterLab installed
- D
AI Platform Notebooks
Show answer and explanation
Correct answer: A
Explanation
Vertex AI Workbench (Managed Notebooks) is a fully managed Jupyter Notebook environment on Google Cloud that supports GPU acceleration, integrates seamlessly with other Vertex AI services, and provides a high level of scalability and customization. This makes it the optimal choice for training large deep learning models requiring GPU resources.
- A. Correct.
Vertex AI Workbench (Managed Notebooks) is the most appropriate option as it supports GPU acceleration, is fully managed, and integrates seamlessly with other Vertex AI services.
- B. Incorrect.
Google Colab is a free and accessible platform, but it is not specifically designed for seamless integration with Vertex AI services, and its GPU resources are not as customizable or scalable as Vertex AI Workbench.
- C. Incorrect.
Using a custom Compute Engine VM with JupyterLab installed can provide GPU acceleration, but it lacks the managed environment and direct integration with Vertex AI services, which adds extra operational overhead.
- D. Incorrect.
AI Platform Notebooks was a precursor to Vertex AI Workbench and is now deprecated in favor of Vertex AI Workbench. Thus, it is not the recommended choice.