4.7.1.1 The Model Catalog Table

The MODEL_CATALOG table (ML_SCHEMA_user_name.MODEL_CATALOG) has the following columns:

• model_id

  A primary key, and a unique auto-incrementing numeric identifier for the model.

• model_handle

  A name for the model. The model handle must be unique in the model catalog. The model handle is generated or set by the user when the ML_TRAIN routine runs on a training dataset. The generated model_handle format is schemaName_tableName_userName_No, as in the following example: heatwaveml_bench.census_train_user1_1636729526. See Work with Model Handles to learn more.

  Note

  The format of the generated model handle is subject to change.

• model_object

  Set to null. Models are stored in the model_object_catalog table.

• model_owner

  The user who initiated the ML_TRAIN query to create the model.

• build_timestamp

  A timestamp indicating when the model was created (in UNIX epoch time). A model is created when the ML_TRAIN routine finishes running.

• target_column_name

  The name of the column in the training table that was specified as the target column.

• train_table_name

  The name of the input table specified in the ML_TRAIN query.

• model_object_size

  The model object size, in bytes.

• model_type

  The type of model (algorithm) selected by ML_TRAIN to build the model.

• task

  The task type specified in the ML_TRAIN query.

• column_names

  The feature columns used to train the model.

• model_explanation

  The model explanation generated during training. See Generate Model Explanations.

• last_accessed

  The last time the model was accessed. AutoML routines update this value to the current timestamp when accessing the model.

• model_metadata

  Metadata for the model. If an error occurs during training or you cancel the training operation, AutoML records the error status in this column. See Model Metadata.

• notes

  Use this column to record notes about the trained model. It also records any error messages that occur during model training.

The Model Object Catalog Table

Models are chunked and stored uncompressed in the model_object_catalog table. Each chunk is saved with the same model_handle.

A call to one of the following routines upgrades the model catalog, and store the model in the model_object_catalog table:

• ML_TRAIN

• ML_MODEL_LOAD

• ML_EXPLAIN

• ML_MODEL_IMPORT

• ML_MODEL_EXPORT

If the call to one of these routines is not successful or is aborted, then the previous model catalog is still available.

The model_object_catalog table has the following columns:

• chunk_id

  A primary key, and an auto-incrementing numeric identifier for the chunk. It is unique for the chunks sharing the same model_handle.

• model_handle

  A primary key, and a foreign key that references model_handle in the MODEL_CATALOG table.

• model_object

  A string in JSON format containing the serialized AutoML model.

See Also

• Review Model Metadata for the Model Catalog Table.

• Review Model Handles and how to retrieve them from the Model Catalog Table.