Use Model export to save a model to an OSS path - Platform For AI

You can use the Export General-Purpose Model component to export a model trained in MaxCompute to a specified Object Storage Service (OSS) path.

Prerequisites

The OSS authorization is complete. For more information, see Cloud Product Dependencies and Authorization: Designer.

Supported upstream components

General machine learning components
PS-series components

Visual Configuration Components

On the Parameters tab of the Export General-Purpose Model component, configure the following parameters.

Parameter	Description
Rename	If this parameter is configured, the exported model is renamed before it is saved. If this parameter is not configured, the exported model is saved with the name of the upstream output model. Note To overwrite the previously exported model each time you run the export, configure Rename and select the Overwrite Existing Model parameter.
Overwrite	If this parameter is selected, the generated model file overwrites the model file of the same name in the specified OSS path.
Export Model Format	The following values are supported: original: The exported model includes a `.meta` file and a `.model` file that describe the model. pmml: The exported model file has a `.xml` extension. Model formats that can be exported by upstream components: General machine learning components support exporting models in original and pmml formats. Connect the output model to port 1 of the Export General-Purpose Model component. PS-series components support exporting models only in the original format. The downstream component must be connected to both port 1 and port 2 of the Export General-Purpose Model component.

Configure the component using PAI commands

You can configure the component by running the following PAI command in an SQL script or an ODPS SQL node in DataWorks.

PAI -name generalmodeltransfer2oss
-project algo_public
-Dformat="original"
-Drename="model_export"
-Doverwrite="false"
-DossPath="oss://examplebucket-cn-hangzhou-internal.aliyuncs.com/export/"
-Darn="acs:ram::xxxxxxxxxx:role/aliyunodpspaidefaultrole"
-DmodelName="model_flow_aius5tamq5rv4x****_node_anprs9ufo40opc****_model"

The following table describes the parameters.

Parameter Name	Required	Description	Default Value
format	Yes	The format of the output model. The following values are supported: original pmml	None
rename	No	The new name for the model.	None (The model is not renamed by default.)
overwrite	No	Specifies whether to overwrite a model of the same name in the specified directory. The following values are supported: true false	false
ossPath	Yes	The OSS path to which the model is exported.	None
arn	Yes	The authorization information. To obtain this information, log on to the PAI console. On the All Product Dependencies page, find the Designer section and click View Authorization Information in the Actions column. For more information, see Cloud product dependencies and authorization: Designer.	None
inputTable	No	When you export a model from a PS-series component, you must also specify the model table that is generated by the training component.	None
modelName	Yes	The name of the offline model to export.	None