Qeexo Model Converter
Convert your tree-based machine learning models for the Edge!
existing tree-based machine learning models
to run on Arm Cortex-M0 to Cortex-M4 Platforms
Patent-pending quantization technology for tree-based models
Qeexo Model Converter quantizes existing pre-built models to reduce memory consumption so they can fit onto constrained embedded devices.
Users can customize memory constraints
Need models of specific sizes? Qeexo Model Converter will generate the best-performing model under a given size.
Takes ONNX file format as input; can handle multiple tree formats
Gradient Boosting Machine (GBM), Random Forest (RF), and XGBoost (XGB) models are currently supported. New models (including neural networks) will be added in the near future.
Generates optimized object code for Cortex-M0 to Cortex-M4 platforms
Qeexo Model Converter outputs optimized object code that can readily be integrated into codebase for embedded platforms.
Converter User Guide
First month free, then $30/month for unlimited use.