scikit-learn-intelex
scikit-learn-intelex copied to clipboard
intel
Intel(R) Extension for Scikit-learn is a seamless way to speed up your Scikit-learn application
# Description PCA support was added - internal Python interface was implemented - external Python wrapper for decomposition.PCA was implemented - pyBind11 bindings were provided - small changes for correct...
I use sklearnex to patch original sklearn. In my application I call KMeans from sklearn.cluster >50 times. And I have the next message every time: `[INFO] sklearn.cluster.KMeans.fit: uses Intel(R) oneAPI...
**Describe the bug** I run the simplest example of daal4py. It outputs ModuleNotFoundError: No module named 'daal4py._oneapi' **To Reproduce** Steps to reproduce the behavior: reproducer ```python from sklearnex import patch_sklearn,...
The daal4py `kmeans` function accepts an argument `maxIterations`. It is implied that this will bound the _maximum_ number of updates that the algorithm will undergo in an attempt to find...
**Describe the bug** scikit-learn-intelex Random Forest hangs on KDDCup09 dataset (Large binary classification 2GB dataset in AutoMLBenchmark) - 45000 rows - 1260 float columns - 6223 int columns - 5797...
Version used: 2021.5.1 The following model gets a speedup with `predict_proba`: ``` from sklearnex.ensemble import RandomForestClassifier model = RandomForestClassifier(..., criterion="gini") ``` However, this does not get a speedup: ``` from...
The development version of scikit-learn now uses a unified loss module for several estimators including Logistic Regression: https://github.com/scikit-learn/scikit-learn/pull/21808 This causes the following problem when using the latest version of scikit-learn-intelex...
**Describe the bug** CC: @mfeurer When using `python 3.10`, pip says that `daal4py` version `daal4py-2021.5.3-py39-none-manylinux1_x86_64.whl` gets downloaded, the incorrect `py` version. This causes an import error [here](https://github.com/intel/scikit-learn-intelex/blob/master/daal4py/__init__.py#L30). This came about...
