Question 18
Domain 3You are training an LSTM-based model on Vertex AI to summarize text using the following job submission script: ``` gcloud ai-platform jobs submit training $JOB_NAME \ --package-path $TRAINER_PACKAGE_PATH \ --module-name $MAIN_TRAINER_MODULE \ --job-dir $JOB_DIR \ --region $REGION \ --scale-tier basic \ -- \ --epochs 20 \ --batch_size=32 \ --learning_rate=0.001 ``` You want to ensure that training time is minimized without significantly compromising the accuracy of your model. What should you do?
Correct answer: B
Explanation
The "scale-tier" parameter controls the training infrastructure size, so increasing it can reduce training time by using more compute resources. Since the goal is to "minimize" training time without a major accuracy tradeoff, changing the scale tier is the lever that affects speed rather than model hyperparameters like "epochs" or "learning_rate".
Why each option is right or wrong
A. Modify the ‘epochs’ parameter.
B. Modify the 'scale-tier' parameter.
In the `gcloud ai-platform jobs submit training` command, `--scale-tier` is the field that selects the amount of training infrastructure allocated by Vertex AI/AI Platform, so it directly affects wall-clock training time rather than the learned model itself. The `basic` tier is the smallest preset; moving to a larger tier increases available compute and can shorten training without changing the model’s hyperparameters such as `--epochs 20`, `--batch_size=32`, or `--learning_rate=0.001`, which would alter optimization behavior and could affect accuracy.
C. Modify the 'batch size' parameter.
D. Modify the 'learning rate' parameter.