Generate External Model | BYOM | Teradata Vantage - 3.0 - Generating External Models for Scoring in BYOM - Teradata Vantage

Teradata Vantageā„¢ - Bring Your Own Model User Guide

Teradata Vantage
Release Number
May 2022
Last Update
Content Type
User Guide
Publication ID
English (United States)


Vantage BYOM PMMLPredict function supports the following external models:
  • Anomaly Detection
  • Association Rules
  • Cluster
  • General Regression
  • k-Nearest Neighbors
  • Naive Bayes
  • Neural Network
  • Regression
  • Ruleset
  • Scorecard
  • Random Forest
  • Decision Tree
  • Vector Machine
  • Multiple Models

Some Machine Learning tools and platforms can generate PMML models directly. Other models must be converted to PMML. For examples of Python pipelines that generate PMML models, see Python Pipelines for PMML Models.


Vantage BYOM H2OPredict supports Driverless AI and H2O-3 MOJO models.

For examples of Python pipelines that generate MOJO models, see Python Code to Generate H2O Open Source Models.


Vantage BYOM ONNXPredict supports models in ONNX format. Several training frameworks support native export functionality to ONNX, such as Chainer, Caffee2, and PyTorch. You can also convert models from several toolkits (such as, scikit-learn, TensorFlow, Keras, XGBoost, H2O, and Spark ML) to ONNX.

For maximum model sizes supported by BYOM, see Maximum Model Size.