Xavier Dupré

PR #19591 fixes this issue. Am I allowed to use the model to write the unit test checking the bug never shows up again?

This error message lacks precision. It is returned when the following function fails: https://github.com/microsoft/onnxruntime/blob/main/onnxruntime/core/graph/graph_utils.cc#L541. The reason why a node cannot be replaced is one of the reasons mentioned in the...

It seems minimal_build and shared_lib are not supported altogether (https://github.com/microsoft/onnxruntime/blob/main/cmake/onnxruntime_providers_cpu.cmake#L220). +@skottmckay

The source are provided by default when github publishes a new releases but git dependencies are not included. To build a specific version, you need to clone the reposity and...

It is still needed? There should be a unit test to check this works.

How many features do you use? onnxruntime does not support sparse tensor for tree ensemble right now, only dense. That might explain the memory explosion.

I guess that's the reason. Is sparse needed in your case or can you do without?

I'm closing the issue, feel free to reopen if needed.

Closing the issue, sklearn-onnx will use [onnx-script](https://github.com/microsoft/onnx-script) next year.

I would do this in a different way. I think we should export in python all functions doing shape inference for every operator. We should have a python version of...