ONNX benchmarks¶
The following benchmarks compare runtime or backeend with ONNX.
One-Off predictions¶
The following benchmark measures the prediction time between scikit-learn and onnxruntime for different configurations related to one-off predictions: predictions are computed for one observation at a time which is the standard scenario in a webservice. onnxruntime allows for some models to run batch predictions. If this functionality is available, it is usually tested for small batches (like 10 observations).
- Prediction time scikit-learn / onnxruntime: summary
- Prediction time scikit-learn / onnxruntime for LogisticRegression
- Prediction time scikit-learn / onnxruntime for MultinomialNB
- Prediction time scikit-learn / onnxruntime for DecisionTreeClassifier
- Prediction time scikit-learn / onnxruntime for RandomForestClassifier