Note
Access to this page requires authorization. You can try signing in or changing directories.
Access to this page requires authorization. You can try changing directories.
Note
This information applies to Databricks CLI versions 0.205 and above. The Databricks CLI is in Public Preview.
Databricks CLI use is subject to the Databricks License and Databricks Privacy Notice, including any Usage Data provisions.
The model-versions command group within the Databricks CLI contains commands to manage model versions in the model registry in Unity Catalog. Models in Unity Catalog provide centralized access control, auditing, lineage, and discovery of ML models across Databricks workspaces. See Manage model lifecycle in Unity Catalog.
databricks model-versions delete
Delete a model version from the specified registered model. Any aliases assigned to the model version will also be deleted.
The caller must be a metastore admin or an owner of the parent registered model. For the latter case, the caller must also be the owner or have the USE_CATALOG privilege on the parent catalog and the USE_SCHEMA privilege on the parent schema.
databricks model-versions delete FULL_NAME VERSION [flags]
Arguments
FULL_NAME
The three-level (fully qualified) name of the model version
VERSION
The integer version number of the model version
Options
Examples
The following example deletes version 1 of a model:
databricks model-versions delete main.my_schema.my_model 1
databricks model-versions get
Get a model version.
The caller must be a metastore admin or an owner of (or have the EXECUTE privilege on) the parent registered model. For the latter case, the caller must also be the owner or have the USE_CATALOG privilege on the parent catalog and the USE_SCHEMA privilege on the parent schema.
databricks model-versions get FULL_NAME VERSION [flags]
Arguments
FULL_NAME
The three-level (fully qualified) name of the model version
VERSION
The integer version number of the model version
Options
--include-aliases
Whether to include aliases associated with the model version in the response.
--include-browse
Whether to include model versions in the response for which the principal can only access selective metadata for.
Examples
The following example gets version 1 of a model:
databricks model-versions get main.my_schema.my_model 1
The following example gets version 1 of a model with aliases included:
databricks model-versions get main.my_schema.my_model 1 --include-aliases
databricks model-versions get-by-alias
Get a model version by alias.
The caller must be a metastore admin or an owner of (or have the EXECUTE privilege on) the registered model. For the latter case, the caller must also be the owner or have the USE_CATALOG privilege on the parent catalog and the USE_SCHEMA privilege on the parent schema.
databricks model-versions get-by-alias FULL_NAME ALIAS [flags]
Arguments
FULL_NAME
The three-level (fully qualified) name of the registered model
ALIAS
The name of the alias
Options
--include-aliases
Whether to include aliases associated with the model version in the response.
Examples
The following example gets a model version by alias:
databricks model-versions get-by-alias main.my_schema.my_model production
The following example gets a model version by alias with aliases included:
databricks model-versions get-by-alias main.my_schema.my_model production --include-aliases
databricks model-versions list
List model versions. You can list model versions under a particular schema, or list all model versions in the current metastore.
The returned models are filtered based on the privileges of the calling user. For example, the metastore admin is able to list all the model versions. A regular user needs to be the owner or have the EXECUTE privilege on the parent registered model to recieve the model versions in the response. For the latter case, the caller must also be the owner or have the USE_CATALOG privilege on the parent catalog and the USE_SCHEMA privilege on the parent schema.
There is no guarantee of a specific ordering of the elements in the response. The elements in the response will not contain any aliases or tags.
databricks model-versions list FULL_NAME [flags]
Arguments
FULL_NAME
The full three-level name of the registered model under which to list model versions
Options
--include-browse
Whether to include model versions in the response for which the principal can only access selective metadata for.
--max-results int
Maximum number of model versions to return.
--page-token string
Opaque pagination token to go to next page based on previous query.
Examples
The following example lists all model versions for a registered model:
databricks model-versions list main.my_schema.my_model
The following example lists model versions with pagination:
databricks model-versions list main.my_schema.my_model --max-results 10 --page-token abc123token
databricks model-versions update
Update the specified model version.
The caller must be a metastore admin or an owner of the parent registered model. For the latter case, the caller must also be the owner or have the USE_CATALOG privilege on the parent catalog and the USE_SCHEMA privilege on the parent schema.
Currently only the comment of the model version can be updated.
databricks model-versions update FULL_NAME VERSION [flags]
Arguments
FULL_NAME
The three-level (fully qualified) name of the model version
VERSION
The integer version number of the model version
Options
--comment string
The comment attached to the model version.
--json JSON
The inline JSON string or the @path to the JSON file with the request body
Examples
The following example updates a model version comment:
databricks model-versions update main.my_schema.my_model 1 --comment "Updated model with improved accuracy"
The following example updates a model version using a JSON file:
databricks model-versions update main.my_schema.my_model 1 --json @model-version-update.json
Global flags
--debug
Whether to enable debug logging.
-h or --help
Display help for the Databricks CLI or the related command group or the related command.
--log-file string
A string representing the file to write output logs to. If this flag is not specified then the default is to write output logs to stderr.
--log-format format
The log format type, text or json. The default value is text.
--log-level string
A string representing the log format level. If not specified then the log format level is disabled.
-o, --output type
The command output type, text or json. The default value is text.
-p, --profile string
The name of the profile in the ~/.databrickscfg file to use to run the command. If this flag is not specified then if it exists, the profile named DEFAULT is used.
--progress-format format
The format to display progress logs: default, append, inplace, or json
-t, --target string
If applicable, the bundle target to use