Serving JPMML-based Tree-based models on Amazon SageMaker

This example notebook demonstrates how to bring your own java-based container and serve PMML based Models on Amazon SageMaker for Inference.

Getting Started

• Open AWS Console
• Search “SageMaker” and click “Amazon Sagemaker”

• Click Notebook -> Notebook Instances -> Create notebook Instances

• Create Notebook Instance with the following details

  Notebook instance name – “sagemaker-byoc-mme-java”

  Notebook instance type – “ml.m5.xlarge”

• Go back to Notebook instance tab page and Open the notebook instance when status is InService -> Click “Open JupyterLab”. This will open a JupyterLab Interface.

• Go to "Git" from the top menu and select "Clone a Repository" as shown below

- Clone this repo - https://github.com/aws/amazon-sagemaker-examples.git as shown below

• Navigate to inference->realtime->byoc->byoc-mme-java folder -> JPMML_Models_SageMaker.ipynb and follow the instructions

• Select conda_python3 as the kernel in order to execute the notebook.