Unlock the power of your data in the cloud! Get hands-on with Google Cloud's core data services like BigQuery and Looker to validate your practical skills in data ingestion, analysis, and management, and earn your Associate Data Practitioner certification!
Prepare and test your skills
Prepare and test your skills
Worked example. The correct answer is already marked and every option is explained below, so there is nothing to select here. To answer questions yourself, start the free trial.
Keep the momentum going with these hand-picked practice scenarios
Want more questions like this?
Get a free certification question every week.
Last updated
An organization manages multiple versions of a fraud detection machine learning model in Model Registry. The engineering team wants to streamline the model release process so that downstream serving applications can automatically consume newly promoted models (such as staging or production) without needing application code or configuration updates whenever a new version ID is generated.
Which approach should the team implement in Model Registry?
Create separate Model Registry parent resources for staging and production and copy artifacts between them during promotion.
Delete the previous model version and upload newly trained model artifacts under the exact same immutable version ID.
Apply resource labels with stage metadata to each model version and configure downstream applications to query the registry by filtering on label values.
Assign a mutable model alias to the target model version and configure downstream applications to reference the model using the alias name.
Create separate Model Registry parent resources for staging and production and copy artifacts between them during promotion.
Delete the previous model version and upload newly trained model artifacts under the exact same immutable version ID.
Apply resource labels with stage metadata to each model version and configure downstream applications to query the registry by filtering on label values.
Assign a mutable model alias to the target model version and configure downstream applications to reference the model using the alias name.
A model alias in Model Registry is a mutable, named pointer assigned to a specific version of a machine learning model. Similar to Git branch references or Docker tags (such as :latest or :stable), an alias allows users and automated systems to identify, deploy, or query a model version using a human-readable text string rather than an immutable, generated version ID.
StudyCerts
StudyCertsstaging or production can be assigned directly to specific model versions to clearly demarcate their lifecycle stages.fraud_detector@production). When a new model version is validated, moving the alias to the new version instantly redirects application traffic without modifying serving code.default alias, ensuring standard calls without explicit versions function smoothly.Using model version aliases decouples model deployment and serving infrastructure from model retraining pipelines. Data science teams can train, test, and promote models independently while client applications consistently query stable alias endpoints.