Closed DanieleMorotti closed 1 year ago
Thanks for the question. OMLT unfortunately does not currently support multi-output GBT models, so it looks like your model is correctly failing the assertion error of having more than one output. We do provide support for multi-output neural network models as you noted. Please see some previous discussion on potential ways to extend this here: https://github.com/cog-imperial/OMLT/discussions/90
It seems that OMLT doesn't work when using lightgbm with a multiclass classification problem (4 classes). I initialized the lightgbm instance as:
and I used the same code that you can find in
docs/notebooks/bo_with_trees.ipynb
to get the ONNX model and to build the tree, it returns the following error:that it seems to be related to the number of classes. I solved the problem using a simple PyTorch neural network and OMLT works well in that case, but I was wondering if I did something wrong using the lightgbm model.
The version of lightgbm is 3.3.5 and the omlt version is 1.1.
Thanks for your attention.