The EXPORT MODEL statement

Overview

To export an existing model from BigQuery ML to Cloud Storage, use the EXPORT MODEL statement.

For more information about supported model types, formats, and limitations, see Exporting models.

For information about supported model types of each SQL statement and function, and all supported SQL statements and functions for each model type, read End-to-end user journey for each model.

Syntax

The following is the syntax of EXPORT MODEL for a regular model that is not generated from BigQuery ML hyperparameter tuning.

EXPORT MODEL MODEL_NAME [OPTIONS(URI = STRING_VALUE)]
  • MODEL_NAME is the name of the BigQuery ML model you're exporting. If you are exporting a model in another project, you must specify the project, dataset, and model in the following format, including backticks:

    `PROJECT.DATASET.MODEL`
    

    For example, `myproject.mydataset.mymodel`.

    If the model name does not exist in the dataset, the following error is returned:

    Error: Not found: Model myproject:mydataset.mymodel

  • STRING_VALUE is the URI of a Cloud Storage bucket where the model is exported. This option is required for the EXPORT MODEL statement. For example:

    URI = 'gs://bucket/path/to/saved_model/'
    

For a model that is generated from BigQuery ML hyperparameter tuning, EXPORT MODEL can also export an individual trial to a destination URI. For example:

EXPORT MODEL MODEL_NAME [OPTIONS(URI = STRING_VALUE [, TRIAL_ID = INT_VALUE])]
  • INT_VALUE is the numeric ID of the exporting trial. For example:

    ```sql
    TRIAL_ID = 12
    ```
    
  • If TRIAL_ID is not specified, then the optimal trial is exported by default.