Xavier Dupré

You could use hummingbird to convert your tree into a neural network and then run it on GPU.

Could you use this example as a reference https://onnx.ai/sklearn-onnx/auto_tutorial/plot_gexternal_lightgbm.html ?

This case is not exposed as an example but is tested in a unit test: https://github.com/onnx/sklearn-onnx/blob/main/tests_onnxmltools/test_lightgbm.py. You should find the missing pieces in that file.

I tried the following example and it worked. This issue comes from a feature your tree is using and which is not supported by the converter. Given the error you...

It is difficult without knowing the tree. I don't have a way to replicate. However, if you are willing to write some code, you probably modify onnxmltools in the lightgbm...

Thanks for sharing your model. Lightgbm create list with mixed integers and floats for nodes values and ONNX checks there is only one type. I created a PR to fix...

I guess inpsect expects to see physical files but pyinstaller removes the physical file. It could happen with other modules. The code using inpsect is just doing validation and could...

The code with AdaBoostRegressor works with the latest version of skl2onnx. I suggest updating the module.

onnxmltools or sklearn-onnx do not support this library. The prediction function must be implemented with onnx primitives in a converter. This tutorial shows what it means: https://onnx.ai/sklearn-onnx/auto_tutorial/plot_icustom_converter.html.

The supported version for each converting library is wrong. Maybe, we should remove it from this page. The latest version is usually supported.