Xavier Dupré

I could not replicate your issue. I updated your example to make it work. The main issues were the fact the model was not trained before being converted and the...

onnxmltools is still tested against coremltools==3.1 and fails with more recent version. We cannot give any date when supporting more recent versions happens.

Sorry for the delay. I'll look into that this week.

ONNX does not have an official operator to tokenize strings. One custom operator is implemented by onnxruntime and uses re2 to split a string into words. This option is supported...

For historical reason, the converter adds an operator ZipMap by default to represent probabilities. To remove it and get a matrix of probabilities instead of a zipmap operator, argument `options={'zipmap':...

See notebook [Benchmark Random Forests, Tree Ensemble](http://www.xavierdupre.fr/app/mlprodict/helpsphinx/gyexamples/plot_time_tree_ensemble.html#sphx-glr-gyexamples-plot-time-tree-ensemble-py).

Is it possible to know the base estimator used?

Could you try again with the latest package and copy paste the call stack if the error is still happening?

Hi, I would do it that way: ``` # **Imports** import numpy as np from sklearn.base import BaseEstimator, TransformerMixin from sklearn.pipeline import make_pipeline from onnxruntime import InferenceSession from skl2onnx import...

It depends on your model. I think it is better to keep as much as as possible inside onnxruntime. There is room for improvments in the implementation of Where in...