Currently we pickle the models to save the trained model state, which is unsatisfactory. there is now a scikit-learn add on called sklearn-onnx which allows us to save th classifier state in the ONNX format, which is (becoming?) the standard format for interoperability.
https://onnx.ai/sklearn-onnx/supported.html
Currently we pickle the models to save the trained model state, which is unsatisfactory. there is now a scikit-learn add on called sklearn-onnx which allows us to save th classifier state in the ONNX format, which is (becoming?) the standard format for interoperability. https://onnx.ai/sklearn-onnx/supported.html