Description
Deploying a trained XGBoost model drags along a whole Python environment with slow inference. Timber is Ollama for classical ML models, compiling them into native C99 inference code.
It supports XGBoost, LightGBM, scikit-learn, CatBoost and ONNX, one command to load and serve, hundreds of times faster than Python.
AOT:C99 output.
Frameworks:XGBoost to ONNX.
Fast:Hundreds of times.
Serve:One command.
It supports XGBoost, LightGBM, scikit-learn, CatBoost and ONNX, one command to load and serve, hundreds of times faster than Python.
Features
AOT:C99 output.
Frameworks:XGBoost to ONNX.
Fast:Hundreds of times.
Serve:One command.
