Syntax

A google_ml_engine_model is used to test a Google Model resource

Examples

describe google_ml_engine_model(project: 'chef-gcp-inspec', name: 'ml_model') do
  it { should exist }
  its('description') { should cmp 'My awesome ML model' }
  its('regions') { should include 'us-central1' }
  its('online_prediction_logging') { should cmp 'true' }
  its('online_prediction_console_logging') { should cmp 'true' }
end

describe google_ml_engine_model(project: 'chef-gcp-inspec', name: 'nonexistent') do
  it { should_not exist }
end

Properties

Properties that can be accessed from the google_ml_engine_model resource:

• name: The name specified for the model.

• description: The description specified for the model when it was created.

• default_version: The default version of the model. This version will be used to handle prediction requests that do not specify a version.

  • name: The name specified for the version when it was created.

• regions: The list of regions where the model is going to be deployed. Currently only one region per model is supported

• online_prediction_logging: If true, online prediction access logs are sent to StackDriver Logging.

• online_prediction_console_logging: If true, online prediction nodes send stderr and stdout streams to Stackdriver Logging

• labels: One or more labels that you can add, to organize your models.

GCP Permissions

Ensure the Cloud ML is enabled for the current project.