Question 23
Domain 3Your data science team is in the process of training a PyTorch model for image classification, building upon a pre-trained ResNet model. To achieve optimal performance, you now find the need to conduct hyperparameter tuning for various parameters. What steps should you take in this scenario?
Correct answer: B
Explanation
Vertex AI supports hyperparameter tuning for custom training jobs, including PyTorch models packaged in custom containers. Running a tuning job lets you search over parameters like learning rate, batch size, and optimizer settings to improve model performance during training.
Why each option is right or wrong
A. Convert the model to a Keras model and run a Keras Tuner job.
B. Run a hyperparameter tuning job on Vertex AI using custom containers.
Vertex AI’s hyperparameter tuning service is invoked as a tuning job for a custom training workload, which is the correct fit when the model is a PyTorch ResNet packaged in a custom container. Under the Vertex AI custom training and hyperparameter tuning workflow, you submit the containerized training job and specify the search space for parameters such as learning rate, batch size, and optimizer settings; Vertex AI then launches multiple trials automatically.
C. Create a Kubeflow Pipelines instance and run a hyperparameter tuning job on Katib.
D. Convert the model to a TensorFlow model and run a hyperparameter tuning job on Vertex AI.