readme.md

Wrapping Your Model

To allow your component (model, router etc) to be managed by seldon-core it needs

1. To be built into a Docker container
2. To expose the approripiate service APIs over REST or gRPC.

To wrap your model follow the instructions for your chosen language or toolkit.

Python

Python based models, including TensorFlow, Keras, pyTorch, StatsModels, XGBoost and Scikit-learn based models.

You can use either:

• Source-to-image (s2i) tool
• Seldon Docker wrapper application

R

• R models can be wrapped using source-to-image

Java

Java based models including, H2O, Deep Learning 4J, Spark (standalone exported models).

• Java models wrapped using source-to-image

H2O

H2O models can be wrapped either from Java or Python.

• Java models wrapped using source-to-image
• H2O models saved and called from python