Alexander Andreev

CPU and GPU configs for XGBoost have only few differences: in `data-format` (pandas vs cudf) and `tree-method` (hist vs gpu_hist). Dispatching for them with `--devices(s)` argument will simplify configs.

`conda list` command raises error if develop versions of daal4py and sklearnex are installed in conda environment: > CondaError: Expected exactly one \`egg-info\` directory in '/user_dir/scikit-learn-intelex', via egg-link 'lib/python3.10/site-packages/daal4py.egg-link'. Instead...

Originates from https://github.com/intel/scikit-learn-intelex/issues/968 Source distribution will make possible to build daal4py/sklearnex locally if package for selected OS/python version is not available, but requires to have build dependencies like compiler.

Fix for 'unused variable' DPC++ compiler error

2

# Description Fix for 'unused variable' DPC++ compiler error when assertion (`ONEDAL_ENABLE_ASSERT`) is disabled.

# Description Update develop branch with two last commits from master

# Description Apply clang-format-14 for cpp source files

Estimator parameters tuning

2

# Description Estimator parameters tuning based on OptunaSearchCV. Currently, prototype for SVC only.

Refactoring of benchmarks

13

# Benchmarks rework **Entry points: [README](https://github.com/Alexsandruss/scikit-learn_bench/tree/dev/refactor/README.md), [Developer Guide](https://github.com/Alexsandruss/scikit-learn_bench/blob/dev/refactor/docs/README.md)** ## Key features: - Benchmarks runner, report generator and separate benchmarks are implemented to run as python modules - Benchmarking config specification...