FIL Backend Examples

This directory contains example notebooks which illustrate typical workflows and use-cases for the Triton FIL backend. Additional examples will be added to this directory over time.

Each subdirectory contains an example notebook and a README with instructions on how to run the example.

Current Examples

• Categorical Fraud Example: This introductory example walks through training a categorical XGBoost model for fraud detection and deploying it on both GPU-accelerated and CPU-only systems.

• FAQ Notebook: This notebook answers a series of frequently asked questions around the FIL backend for Triton and offers example code with practical applications of those answers.

Deprecated Examples

• Simple XGBoost: This example has been superseded by the Categorical Fraud Example, which offers a more succinct and up-to-date example of how to train and deploy an XGBoost model.