Google Professional Machine Learning Engineer Question 171
Select 3Google Cloud PlatformYou are a Machine Learning Engineer tasked with building and training a deep learning model using TensorFlow on Vertex AI Workbench. The model will predict customer churn based on historical data. Which of the following steps are necessary to successfully develop your model in Vertex AI Workbench while leveraging cloud resources efficiently?
- A
Install the TensorFlow library in the Vertex AI Workbench environment if it is not pre-installed.
- B
Select a high-memory, GPU-enabled machine type for the Workbench environment to accelerate training.
- C
Use the Vertex AI Workbench's built-in feature to deploy the model to a BigQuery table for training.
- D
Load your data from a Google Cloud Storage bucket into the Workbench environment for preprocessing and training.
- E
Use pre-configured Jupyter notebooks in Vertex AI Workbench to optimize hyperparameters automatically.
Show answer and explanation
Correct answers: A, B, D
Explanation
Developing models in Vertex AI Workbench involves setting up the environment with necessary libraries, selecting appropriate compute resources, and loading data efficiently. While the Workbench provides many helpful features, some tasks, like hyperparameter tuning, require additional effort or tools. Understanding these steps ensures the efficient use of Vertex AI Workbench for machine learning tasks.
- A. Correct.
TensorFlow might not always be pre-installed in the environment, so you may need to manually install it using a package manager like pip.
- B. Correct.
GPU-enabled machine types can significantly accelerate the training of deep learning models, especially for large datasets.
- C. Incorrect.
BigQuery is a data warehouse and cannot be used to directly deploy a model. Deployment typically involves exporting the model to a format suitable for serving.
- D. Correct.
Loading data from a Google Cloud Storage bucket is a standard and recommended practice for accessing large datasets in cloud-based environments like Vertex AI Workbench.
- E. Incorrect.
While Vertex AI Workbench provides useful pre-configured notebooks, it does not automatically optimize hyperparameters for you. Hyperparameter tuning requires specific tools like Vertex AI Vizier or manual configuration.