Register and call remote AI models using model endpoint management

This page describes how to invoke predictions or generate embeddings using a model, and then register the model endpoint with model endpoint management.

For more information about the google_ml.create_model() function, see Model endpoint management reference.

Before you begin

  • If your model endpoint requires authentication, then enable the google_ml_integration extension.

  • Based on the model provider, set up authentication.

  • Make sure that you use the postgres default username to access your database.

Enable the extension

  1. Set the google_ml_integration.enable_model_support database flag to on for your instance. For more information about setting database flags, see Configure database flags.

  2. Connect to your primary instance using either a psql client or Cloud SQL Studio.

  3. Run the following command to ensure that the google_ml_integration extension is updated to version 1.4.2:

        ALTER EXTENSION google_ml_integration UPDATE TO '1.4.2'

  4. Add the google_ml_integration version 1.4.2 extension using psql:

      CREATE EXTENSION google_ml_integration VERSION '1.4.2';

  5. Optional: Grant permission to a non-super PostgreSQL user to manage model metadata:

      GRANT SELECT, INSERT, UPDATE, DELETE ON ALL TABLES IN SCHEMA google_ml TO NON_SUPER_USER;

    Replace NON_SUPER_USER with the non-super PostgreSQL username.

Set up authentication

The following sections show how to set up authentication before adding a Vertex AI model endpoint or model endpoints hosted within Google Cloud.

Set up authentication for Vertex AI

To use the Google Vertex AI model endpoints, you must add Vertex AI permissions to the IAM-based Cloud SQL service account you use to connect to the database. For more information about integrating with Vertex AI, see Integrate Cloud SQL with Vertex AI.

Set up authentication for custom-hosted models

This section explains how to set up authentication if you're using Secret Manager. For all models except Vertex AI model endpoints, you can store your API keys or bearer tokens in Secret Manager.

If your model endpoint doesn't handle authentication through Secret Manager, then this section is optional. For example, if your model endpoint uses HTTP headers to pass authentication information or doesn't use authentication at all, then don't complete the steps in this section.

To create and use an API key or a bearer token, complete the following steps:

  1. Create a secret in Secret Manager. For more information, see Create a secret and access a secret version.

    The secret name and the secret path are used in the google_ml.create_sm_secret() SQL function.

  2. Grant permissions to the Cloud SQL instance to access the secret.

      gcloud secrets add-iam-policy-binding SECRET_ID \
      --member="serviceAccount:SERVICE_ACCOUNT_EMAIL" \
      --role="roles/secretmanager.secretAccessor"

    Replace the following:

    • SECRET_ID: the secret ID in Secret Manager.
    • SERVICE_ACCOUNT_EMAIL: the email address of the IAM-based Cloud SQL service account. To find this email address, use the gcloud sql instances describe INSTANCE_NAME command and replace INSTANCE_NAME with the name of the instance. The value that appears next to the serviceAccountEmailAddress parameter is the email address.

Text embedding models with built-in support

This section shows how to register model endpoints for model endpoint management.

Vertex AI embedding models

Model endpoint management provides built-in support for all versions of the text-embedding-gecko model by Vertex AI. Use the qualified name to set the model version to either textembedding-gecko@001 or textembedding-gecko@002.

Because both the textembedding-gecko and textembedding-gecko@001 model endpoint IDs are pre-registered with model endpoint management, you can use either of them directly as the model ID. For these models, the extension automatically sets up default transform functions.

Ensure that both the Cloud SQL instance and the Vertex AI model that you're querying are in the same region.

To register the textembedding-gecko@002 model endpoint, call the create_model function:

  CALL
    google_ml.create_model(
      model_id => 'textembedding-gecko@002',
      model_provider => 'google',
      model_qualified_name => 'textembedding-gecko@002',
      model_type => 'text_embedding',
      model_auth_type => 'cloudsql_service_agent_iam');

Custom-hosted text embedding models

This section shows how to register custom model endpoints hosted in networks within Google Cloud.

Adding custom-hosted text embedding model endpoints involves creating transform functions, and optionally, custom HTTP headers. On the other hand, adding custom-hosted generic model endpoints involves optionally generating custom HTTP headers and setting the model request URL.

The following example adds the custom-embedding-model text embedding model endpoint hosted by Cymbal, which is hosted within Google Cloud. The cymbal_text_input_transform and cymbal_text_output_transform transform functions are used to transform the input and output format of the model to the input and output format of the prediction function.

To register custom-hosted text embedding model endpoints, complete the following steps:

  1. Call the secret stored in the Secret Manager:

    CALL
  google_ml.create_sm_secret(
    secret_id => 'SECRET_ID',
    secret_path => 'projects/project-id/secrets/SECRET_MANAGER_SECRET_ID/versions/VERSION_NUMBER');

    Replace the following:

    • SECRET_ID: the secret ID that you set and is subsequently used when registering a model endpoint—for example, key1.
    • SECRET_MANAGER_SECRET_ID: the secret ID set in Secret Manager when you created the secret.
    • PROJECT_ID: the ID of your Google Cloud project.
    • VERSION_NUMBER: the version number of the secret ID.

  2. Create the input and output transform functions based on the following signature for the prediction function for text embedding model endpoints. For more information about how to create transform functions, see Transform functions example.

    The following are example transform functions that are specific to the custom-embedding-model text embedding model endpoint:

    -- Input Transform Function corresponding to the custom model endpoint
CREATE OR REPLACE FUNCTION cymbal_text_input_transform(model_id VARCHAR(100), input_text TEXT)
RETURNS JSON
LANGUAGE plpgsql
AS $$
DECLARE
  transformed_input JSON;
  model_qualified_name TEXT;
BEGIN
  SELECT json_build_object('prompt', json_build_array(input_text))::JSON INTO transformed_input;
  RETURN transformed_input;
END;
$$;
-- Output Transform Function corresponding to the custom model endpoint
CREATE OR REPLACE FUNCTION cymbal_text_output_transform(model_id VARCHAR(100), response_json JSON)
RETURNS REAL[]
LANGUAGE plpgsql
AS $$
DECLARE
  transformed_output REAL[];
BEGIN
  SELECT ARRAY(SELECT json_array_elements_text(response_json->0)) INTO transformed_output;
  RETURN transformed_output;
END;
$$;

  3. Call the create model function to register the custom embedding model endpoint:

    CALL
  google_ml.create_model(
    model_id => 'MODEL_ID',
    model_request_url => 'REQUEST_URL',
    model_provider => 'custom',
    model_type => 'text_embedding',
    model_auth_type => 'secret_manager',
    model_auth_id => 'SECRET_ID',
    model_qualified_name => 'MODEL_QUALIFIED_NAME',
    model_in_transform_fn => 'cymbal_text_input_transform',
    model_out_transform_fn => 'cymbal_text_output_transform');

    Replace the following:

    • MODEL_ID: required. A unique ID for the model endpoint that you define (for example, custom-embedding-model). This model ID is referenced for metadata that the model endpoint needs to generate embeddings or invoke predictions.
    • REQUEST_URL: required. The model-specific endpoint when adding custom text embedding and generic model endpoints—for example, https://cymbal.com/models/text/embeddings/v1. Ensure that the model endpoint is accessible through an internal IP address. Model endpoint management doesn't support external IP addresses.
    • MODEL_QUALIFIED_NAME: required if your model endpoint uses a qualified name. The fully qualified name in case the model endpoint has multiple versions.
    • SECRET_ID: the secret ID you used earlier in the google_ml.create_sm_secret() procedure.

Generic models

This section shows how to register a generic gemini-pro model endpoint from Vertex AI Model Garden, which doesn't have built-in support. You can register any generic model endpoint that is hosted within Google Cloud.

Cloud SQL only supports model endpoints that are available through Vertex AI Model Garden and model endpoints hosted in networks within Google Cloud.

Gemini model

The following example adds the gemini-1.0-pro model endpoint from the Vertex AI Model Garden.

To register the gemini-1.0-pro model endpoint, call the create model function:

```sql
CALL
  google_ml.create_model(
    model_id => 'MODEL_ID',
    model_request_url => 'https://us-central1-aiplatform.googleapis.com/v1/projects/PROJECT_ID/locations/us-central1/publishers/google/models/gemini-1.0-pro:streamGenerateContent',
    model_provider => 'google',
    model_auth_type => 'cloudsql_service_agent_iam');
```

Replace the following:
* <code><var>MODEL_ID</var></code>: a unique ID for the model endpoint that you define (for example,<br> `gemini-1`). This model ID is referenced for metadata that the model endpoint needs to generate embeddings or invoke predictions.
* <code><var>PROJECT_ID</var></code>: the ID of your Google Cloud project.

For more information, see how to invoke predictions for generic model endpoints.

What's next