View BigQuery ML model evaluations

The Model Registry includes tooling to view and compare model versions and evaluate model performance. When you register a BigQuery ML model to the Model Registry, you can view evaluations for binary classification or regression metrics to help you determine the performance of your models. These metrics are automatically synced to Model Registry from BigQuery ML.

To learn more about evaluations in Vertex AI, see Model evaluation in Vertex AI.

To learn more about BigQuery ML evaluate metrics, see BigQuery ML model evaluation overview.

In addition to model evaluations, you can also enable Explainable AI on your BigQuery ML models in Model Registry. To enable a model to get feature-based explanations, see Configure feature-based explanations.

Prerequisites

To get started, you need to follow the prerequisites in the Model Registry model evaluation documentation.

View BigQuery ML model evaluations

Console

  1. In the Google Cloud console, go to the Vertex AI Models page.

  2. Navigate to the BigQuery ML model version.

  3. View metrics in the Evaluate tab.

Compare evaluation metrics

You can only compare models of the same type, BigQuery ML this includes classification and regression metrics. To compare different models, they must be the same type. To learn more about comparing your BigQuery ML models, see Compare evaluation metrics.

What's next