Frameworks#
Before you start, you should take a look at the table below in order to check if the framework you are using is already supported by this package.
Framework |
Converter |
Raw Example |
Supported |
---|---|---|---|
TensorFlow |
tf2onnx |
Yes |
|
Keras |
tf2onnx |
Yes |
|
Scikit-learn |
skl2onnx |
Yes |
|
XGBoost |
Native |
Yes |
|
LightGBM |
Native |
Yes |
|
CoreML |
coremltools |
Yes |
|
CatBoost |
Native |
Yes |
|
Spark |
Native |
Yes |
|
H2O |
Native |
Yes |
Backend#
It is important to note that this package relies on ONNXMLTools, if you want to add a new framework, you should check if it is supported by ONNXMLTools
, and if it is not, you may contribute to their package first.