Deploy a single Designer model as an online service on EAS - Platform For AI

Prerequisites

Train and validate models in Designer before deployment. See Build and debug a model.

One-click deployment

Supported components

Components listed below support one-click deployment to EAS. Other components require manual deployment. See Manual deployment.

Component name	Deployable model format	Matching EAS processor	Configuration
Logistic Regression for Binary Classification	PMML	PMML	Before training, open component settings, go to Fields Setting tab, and enable Whether to Generate PMML.
Gradient Boosting Decision Trees (GBDT) Binary Classification	PMML	PMML
Linear SVM	PMML	PMML
Logistic Regression for Multiclass Classification	PMML	PMML
Random Forest	PMML	PMML
Naive Bayes	PMML	PMML
K-means Clustering	PMML	PMML
GBDT Regression	PMML	PMML
Linear Regression	PMML	PMML
Scorecard Training	PMML	PMML
Text Summarization Training	tgz package	EasyNLP	EasyNLP is provided by PAI in a public OSS bucket and configured automatically.
Image Classification Training (torch)	tgz package	EasyCV	EasyCV is provided by PAI in a public OSS bucket and configured automatically.
PyAlink Script	AlinkModel	Alink	See PyAlink Script.
XGBoost Training	XGBoost	XGBoost	See XGBoost Training.

Deploy a model

Go to Visualized Modeling (Designer), select your workspace, and open the workflow containing your trained model.
Click Model List above the workflow canvas. The system automatically detects and displays all deployable models.
Select your model from the list and click Deploy to EAS to open the Create Service page in EAS console.
Configure deployment parameters.

Model File and Processor Type are populated automatically based on model type. For additional options, see Custom deployment.
Click Deploy. Wait until Service Status shows Running.

Manual deployment

Components listed below require manual deployment. After training, use the Export General-purpose Model component to package your model, export it to OSS, and configure EAS deployment manually.

Component name	Deployable model format	Matching EAS processor	Deployment steps
PS-SMART Binary Classification	PS format	PS algorithm	Connect the Export General-purpose Model component downstream of this component.
PS-SMART Multiclass Classification
PS-SMART Regression

After exporting your model to OSS, configure EAS deployment manually. See Custom deployment.

Troubleshooting

Why is a supported node dimmed and cannot be selected?

Open component node settings, go to Fields Setting tab, and enable Whether to Generate PMML. Re-run the component to enable deployment.

Platform For AI:Deploy a single model as an online service