PMLE Collaborating to manage data and models Practice Question

Drag and drop the steps to deploy a trained TensorFlow model to Vertex AI Prediction in the correct order.

Drag or tap steps into the slots.

Steps

Order

1Step 1

2Step 2

3Step 3

4Step 4

5Step 5

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

✓

Export the model, then upload to GCS, then register as a model, then deploy to endpoint, then test

Export the model, upload to GCS, register as a model, deploy to endpoint, then test.

Answer analysis

For each option: why learners choose it and why it is or isn't the right answer here.

✓
Export the model, then upload to GCS, then register as a model, then deploy to endpoint, then test
Why this is correct
This is the correct order because you must first export the trained model to a standard format (e.g., SavedModel), then upload it to Google Cloud Storage for accessibility, register it as a model resource in Vertex AI, deploy it to an endpoint for serving, and finally test the endpoint to ensure it works.
✗
Upload to GCS, then export the model, then register as a model, then deploy to endpoint, then test
Why it's wrong here
This is incorrect because you cannot upload a model to GCS before exporting it; the export step produces the model artifacts that are uploaded.
✗
Export the model, then register as a model, then upload to GCS, then deploy to endpoint, then test
Why it's wrong here
This is incorrect because a model must be stored in GCS before it can be registered in Vertex AI; the registration references the GCS location.
✗
Export the model, then upload to GCS, then deploy to endpoint, then register as a model, then test
Why it's wrong here
This is incorrect because you must register the model before deploying it; deployment requires a registered model resource.

Go deeper