GBDTMO crash in Google colab (see simple example)

[AlexiaJM]

https://colab.research.google.com/drive/1m_QXVbHbeoho5tF9TUYq6R-PSz3536EU?usp=sharing

[zzd1992]

Please try to degrade numpy. When numpy<=1.19.0, it should be fine. When the version is higher, the interface between numpy and c++ has some problems.

[AlexiaJM]

It still crash, see the colab. It now installs numpy==1.18.0 before everything else.

[zzd1992]

Could you show me the system environments? Based on previous feedback, using anaconda leads error for the multiprocessing module. So it's better to use the native python environment.

And how can I see the detailed traceback in your colab?

[AlexiaJM]

I added 'pip list' to show all installed Python packages in the colab and it doesn't seem to have anaconda installed. It's native python.

There's no way to log the error it seems, but I can tell you that it crashes at the 'booster.set_data((x_train, y_train), (x_valid, y_valid))'.

[AlexiaJM]

Outside colab, the error still happens and it returns a Segmentation fault.

[zzd1992]

We don't have enough information to debug. I guess this is caused by the version of libraries. numba and numpy are the main dependencies of this project. Please try numba==0.54.1 and numpy==1.19.0. This works well in my own device.