Google Professional Machine Learning Engineer Question 143
Select 3Google Cloud PlatformYou are a machine learning engineer tasked with prototyping a deep learning model for image classification using a Jupyter notebook on Google Cloud. Which of the following considerations should you prioritize during the prototyping phase?
- A
Ensuring the notebook can be easily migrated to a production environment.
- B
Using a large dataset to maximize model accuracy during prototyping.
- C
Leveraging TensorFlow or PyTorch for rapid experimentation with pre-built libraries.
- D
Configuring autoscaling for the cloud resources used by the notebook.
- E
Tracking experiments and hyperparameters using tools like Vertex AI or MLflow.
Show answer and explanation
Correct answers: A, C, E
Explanation
During the prototyping phase in a Jupyter notebook, it's important to focus on practices that ensure efficiency, reproducibility, and alignment with future production needs. Migratability, using efficient tools like TensorFlow or PyTorch, and experiment tracking are key considerations. Working with large datasets and configuring autoscaling are not typically priorities at this stage.
- A. Correct.
Ensuring the notebook can be easily migrated to a production environment is critical because prototyping should align with production requirements to avoid significant rework later.
- B. Incorrect.
Using a large dataset during prototyping can be inefficient and slow. It’s better to work with a representative subset of the data and scale up later.
- C. Correct.
Leveraging TensorFlow or PyTorch allows you to quickly use pre-built libraries and models, speeding up the prototyping process.
- D. Incorrect.
Configuring autoscaling is more relevant for production environments and not typically a priority during the prototyping phase.
- E. Correct.
Tracking experiments and hyperparameters is an essential practice during prototyping to manage iterative experimentation and ensure reproducibility.