Add GPU benchmarks support to readme #59

vlad-nazarov · 2021-03-30T12:07:30Z

No description provided.

michael-smirnov

LGTM

PetrovKP · 2021-03-30T21:20:29Z

README.md

-|**[train_test_split](https://scikit-learn.org/stable/modules/generated/sklearn.model_selection.train_test_split.html)**|train_test_split|:white_check_mark:|:x:|:white_check_mark:|:x:|
-|**[GradientBoostingClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
-|**[GradientBoostingRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
+| algorithm  | benchmark name | sklearn | sklearn on GPU | daal4py | cuml | xgboost |


My concerns:

What is sklearn on GPU? (It is not private benchmarks)

It is not aligned with the rest of the benchmarks for libraries: sklearn/xgboost/daal4py/cuml

if you want to highlight algorithms on GPU, I would make it as a note at the bottom of the table.

@SmirnovEgorRu , @michael-smirnov What do you think?

My opinion is we should think about how to present GPU support better, especially because the number of supported algorithms will grow in the future. It's not a supplementary thing to place somewhere in the bottom. In my opinion, it should be on the same level as CPU support for sklearn.
We can remove it from this table, but create another section "Support of Intel(R) Extension for scikit-learn" when to describe which benchmarks are supported on CPU and GPU. @PetrovKP what do you think about this?

…bench into dev/vnazarov_update_readme

PetrovKP · 2021-04-14T01:02:27Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#Intel(R)-Extension-for-Scikit-learn*-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.


Not works reference. Maybe?

Suggested change

* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#Intel(R)-Extension-for-Scikit-learn*-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.

* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](## Intel(R)-Extension-for-Scikit-learn*-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.

PetrovKP · 2021-04-14T01:04:11Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#Intel(R)-Extension-for-Scikit-learn*-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.


Suggested change

* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#Intel(R)-Extension-for-Scikit-learn*-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.

* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#Intel(R)-Extension-for-Scikit-learn*-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.

PetrovKP · 2021-04-14T01:10:42Z

README.md

@@ -108,6 +107,9 @@ The configuration of benchmarks allows you to select the frameworks to run, sele
 |**[GradientBoostingClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
 |**[GradientBoostingRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|

+## Intel(R) Extension for Scikit-learn* support
+[Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) provides the ability to run scikit-learn on CPU and GPU with kernels optimized by [oneDAL](https://github.com/oneapi-src/oneDAL). The extension support GPU patching for algorihms: **DBSCAN**, **KMeans**, **LinearRegression**, **LogisticRegression**. You can launch benchmarks with GPU support using [skl_xpu_config.json](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_with_context_config.json) configuration file.


Suggested change

[Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) provides the ability to run scikit-learn on CPU and GPU with kernels optimized by [oneDAL](https://github.com/oneapi-src/oneDAL). The extension support GPU patching for algorihms: **DBSCAN**, **KMeans**, **LinearRegression**, **LogisticRegression**. You can launch benchmarks with GPU support using [skl_xpu_config.json](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_with_context_config.json) configuration file.

[Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) providing drop-in patching speeds up scikit-learn on CPU and GPU.

Scikit-learn benchmarks on GPU with Intel(R) Extension for Scikit-learn support algorithms: **DBSCAN**, **KMeans**, **LinearRegression**, **LogisticRegression**. Example config with GPU support [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_with_context_config.json).

PetrovKP · 2021-04-14T01:12:25Z

configs/README.md

@@ -27,6 +27,7 @@ Refer to the tables below for descriptions of all fields in the configuration fi
 |data-order| array[string] | **REQUIRED**  input data order. Data order: *C* (row-major, default) or *F* (column-major) |
 |dtype| array[string] | **REQUIRED**  input data type. Data type: *float64* (default) or *float32* |
 |check-finitness| array[] | Check finiteness in sklearn input check(disabled by default) |
+|device| array[string] | For scikit-learn only. List of devices to run with sycl context. It can be *None* (without context, default), *cpu*, *gpu* or *host*|


Minor: what is host device for users? Maybe add description? For runner script no such parameter?

Maybe we just remove the host from here and here https://github.com/vlad-nazarov/scikit-learn_bench/blob/dev/vnazarov_update_readme/configs/skl_xpu_config.json#L7?

michael-smirnov · 2021-04-14T15:12:47Z

README.md

@@ -108,6 +107,11 @@ The configuration of benchmarks allows you to select the frameworks to run, sele
 |**[GradientBoostingClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
 |**[GradientBoostingRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|

+## Intel(R) Extension for Scikit-learn* support
+[Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) providing drop-in patching speeds up scikit-learn on CPU and GPU. 


"providing drop-in patching": there is no final decision yet about how XPUs will be supported. Can we use more neutral wordings like "provides optimized functionality of scikit-learn on Intel(R) CPU and GPU" at the moment?

michael-smirnov · 2021-04-14T15:17:14Z

README.md

+## Intel(R) Extension for Scikit-learn* support
+[Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) providing drop-in patching speeds up scikit-learn on CPU and GPU. 
+
+Scikit-learn benchmarks on GPU with Intel(R) Extension for Scikit-learn support algorithms: **DBSCAN**, **KMeans**, **LinearRegression**, **LogisticRegression**. Example config with GPU support [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json).


The following benchmarks support GPU computations with help of Intel(R) Extension for Scikit-learn*:

dbscan

kmeans ...

Example config with GPU support [here]: The configuration file that contains all of these benchmarks can be found [here].
The link for "here" is broken.
Please add an example of command how to run this config on GPU here

michael-smirnov · 2021-04-14T15:23:58Z

configs/README.md

@@ -27,6 +27,7 @@ Refer to the tables below for descriptions of all fields in the configuration fi
 |data-order| array[string] | **REQUIRED**  input data order. Data order: *C* (row-major, default) or *F* (column-major) |
 |dtype| array[string] | **REQUIRED**  input data type. Data type: *float64* (default) or *float32* |
 |check-finitness| array[] | Check finiteness in sklearn input check(disabled by default) |
+|device| array[string] | For scikit-learn only. List of devices to run with sycl context. It can be *None* (without context, default), *cpu*, *gpu* or *host*|


Maybe we just remove the host from here and here https://github.com/vlad-nazarov/scikit-learn_bench/blob/dev/vnazarov_update_readme/configs/skl_xpu_config.json#L7?

michael-smirnov · 2021-04-15T05:36:33Z

README.md

@@ -108,6 +107,17 @@ The configuration of benchmarks allows you to select the frameworks to run, sele
 |**[GradientBoostingClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
 |**[GradientBoostingRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|

+## Intel(R) Extension for Scikit-learn support
+
+By default scikit-learn benchmark launches using [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) on the CPU (use ``no-intel-optimized`` option to run without extention). Some benchmarks have a GPU support:


Not sure this is correct sentence. Maybe change by "The launches of scikit-learn benchmark use ... by default"?

michael-smirnov · 2021-04-15T05:37:22Z

README.md

+* linear
+* log_reg
+
+A configuration file that contains all these benchmarks can be found [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json). You can use this file to run these benchmarks on both CPU and GPU.


Please add the command to run this config and get a report

michael-smirnov · 2021-04-15T05:37:44Z

configs/README.md

@@ -27,6 +27,7 @@ Refer to the tables below for descriptions of all fields in the configuration fi
 |data-order| array[string] | **REQUIRED**  input data order. Data order: *C* (row-major, default) or *F* (column-major) |
 |dtype| array[string] | **REQUIRED**  input data type. Data type: *float64* (default) or *float32* |
 |check-finitness| array[] | Check finiteness in sklearn input check(disabled by default) |
+|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Please reffer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details|


Suggested change

|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Please reffer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details|

|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Please refer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details|

outoftardis · 2021-04-15T10:35:05Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths


Suggested change

* ``configs`` : configuration files paths

* ``--configs``: specify the path to a configuration file.

outoftardis · 2021-04-15T10:36:42Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.
 * ``output-file``: output file name for result benchmarks. Default is `result.json`
 * ``report``: create an Excel report based on benchmarks results. Need library `openpyxl`.
 * ``dummy-run`` : run configuration parser and datasets generation without benchmarks running.


Suggested change

* ``dummy-run`` : run configuration parser and datasets generation without benchmarks running.

* ``--dummy-run``: run configuration parser and dataset generation without benchmarks running.

outoftardis · 2021-04-15T10:37:51Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:


Suggested change

runner options:

Options:

outoftardis · 2021-04-15T10:40:50Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.
 * ``output-file``: output file name for result benchmarks. Default is `result.json`
 * ``report``: create an Excel report based on benchmarks results. Need library `openpyxl`.


Suggested change

* ``report``: create an Excel report based on benchmarks results. Need library `openpyxl`.

* ``--report``: create an Excel report based on benchmark results. The `openpyxl` library is required.

outoftardis · 2021-04-15T10:43:18Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.
 * ``output-file``: output file name for result benchmarks. Default is `result.json`


Suggested change

* ``output-file``: output file name for result benchmarks. Default is `result.json`

* ``--output-file``: output file name for the benchmark result. The default name is `result.json`.

outoftardis · 2021-04-15T10:46:02Z

README.md

@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

 runner options:
 * ``configs`` : configuration files paths
-* ``no-intel-optimized`` : using Scikit-learn without Intel(R) Extension for Scikit-learn*. Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
+* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.


Now available for [scikit-learn benchmarks] -- why do you need this info?

Suggested change

* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.

* ``--no-intel-optimized``: use Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). By default, the runner uses Intel(R) Extension for Scikit-learn.

outoftardis · 2021-04-15T10:47:12Z

configs/README.md

@@ -27,6 +27,7 @@ Refer to the tables below for descriptions of all fields in the configuration fi
 |data-order| array[string] | **REQUIRED**  input data order. Data order: *C* (row-major, default) or *F* (column-major) |
 |dtype| array[string] | **REQUIRED**  input data type. Data type: *float64* (default) or *float32* |
 |check-finitness| array[] | Check finiteness in sklearn input check(disabled by default) |
+|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Please refer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details|


Suggested change

|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Please refer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details|

|device| array[string] | For scikit-learn only. The list of devices to run the benchmarks on. It can be *None* (default, run on CPU without sycl context) or one of the types of sycl devices: *cpu*, *gpu*, *host*. Refer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details.|

outoftardis · 2021-04-15T10:51:39Z

README.md

@@ -108,6 +107,17 @@ The configuration of benchmarks allows you to select the frameworks to run, sele
 |**[GradientBoostingClassifier](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingClassifier.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|
 |**[GradientBoostingRegressor](https://scikit-learn.org/stable/modules/generated/sklearn.ensemble.GradientBoostingRegressor.html)**|gbt|:x:|:x:|:x:|:white_check_mark:|

+## Intel(R) Extension for Scikit-learn support
+
+The runs of Scikit-learn benchmark use [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) on the CPU by default (use ``no-intel-optimized`` option to run without extention). Some benchmarks have a GPU support:


Suggested change

The runs of Scikit-learn benchmark use [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) on the CPU by default (use ``no-intel-optimized`` option to run without extention). Some benchmarks have a GPU support:

When you run scikit-learn benchmarks on CPU, [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) is used by default. Use the ``--no-intel-optimized`` option to run the benchmarks without the extension.

The following benchmarks have a GPU support:

outoftardis · 2021-04-15T10:53:50Z

README.md

+* linear
+* log_reg
+
+A configuration file that contains all these benchmarks can be found [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json). You can use this file to run these benchmarks on both CPU and GPU.


Suggested change

A configuration file that contains all these benchmarks can be found [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json). You can use this file to run these benchmarks on both CPU and GPU.

You may use the [configuration file for these benchmarks](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json) to run them on both CPU and GPU.

Update readme

672c8cc

vlad-nazarov requested a review from outoftardis as a code owner March 30, 2021 12:07

vlad-nazarov requested review from michael-smirnov and PetrovKP March 30, 2021 12:07

vlad-nazarov added the docs documentation and readme update label Mar 30, 2021

michael-smirnov approved these changes Mar 30, 2021

View reviewed changes

outoftardis approved these changes Mar 30, 2021

View reviewed changes

PetrovKP suggested changes Mar 30, 2021

View reviewed changes

Vladislav Nazarov added 2 commits April 13, 2021 23:45

Merge branch 'master' of https://github.com/IntelPython/scikit-learn_…

19bc2e7

…bench into dev/vnazarov_update_readme

Update GPU readme

53b4200

vlad-nazarov requested review from outoftardis and michael-smirnov April 13, 2021 22:18

Add hyperlink

cbd99b1

PetrovKP reviewed Apr 14, 2021

View reviewed changes

PetrovKP self-requested a review April 14, 2021 01:12

Fix comments

becfbde

michael-smirnov reviewed Apr 14, 2021

View reviewed changes

Vladislav Nazarov added 3 commits April 14, 2021 23:48

Update readme

9874958

Fix hyperlink

0fe0ab4

Update comments

9ef18ab

PetrovKP approved these changes Apr 14, 2021

View reviewed changes

Update link

ece84d9

michael-smirnov suggested changes Apr 15, 2021

View reviewed changes

Fix comments

6568e48

outoftardis reviewed Apr 15, 2021

View reviewed changes

Apply comments

9f133ef

michael-smirnov approved these changes Apr 15, 2021

View reviewed changes

vlad-nazarov merged commit c5ddd0e into IntelPython:master Apr 15, 2021

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add GPU benchmarks support to readme #59

Add GPU benchmarks support to readme #59

vlad-nazarov commented Mar 30, 2021

michael-smirnov left a comment

PetrovKP Mar 30, 2021 •

edited

Loading

michael-smirnov Mar 31, 2021

PetrovKP Apr 14, 2021

PetrovKP Apr 14, 2021

PetrovKP Apr 14, 2021

PetrovKP Apr 14, 2021

michael-smirnov Apr 14, 2021

michael-smirnov Apr 14, 2021

michael-smirnov Apr 14, 2021

michael-smirnov Apr 14, 2021

michael-smirnov Apr 14, 2021

michael-smirnov Apr 15, 2021

michael-smirnov Apr 15, 2021

michael-smirnov Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

outoftardis Apr 15, 2021

	* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn](#Intel(R)-Extension-for-Scikit-learn-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.
	* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn](## Intel(R)-Extension-for-Scikit-learn-support). Now avalible for scikit-learn benchmarks. Default starts with using Intel(R) Extension for Scikit-learn*.

	\|device\| array[string] \| For scikit-learn only. The list of devices to run the benchmarks on. It can be None (default, run on CPU without sycl context) or one of the types of sycl devices: cpu, gpu, host. Please reffer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details\|
	\|device\| array[string] \| For scikit-learn only. The list of devices to run the benchmarks on. It can be None (default, run on CPU without sycl context) or one of the types of sycl devices: cpu, gpu, host. Please refer to [SYCL specification](https://www.khronos.org/files/sycl/sycl-2020-reference-guide.pdf) for details\|

		@@ -67,7 +66,7 @@ Run `python runner.py --configs configs/config_example.json [--output-file resul

		runner options:
		* ``configs`` : configuration files paths

	* ``configs`` : configuration files paths
	* ``--configs``: specify the path to a configuration file.

	* ``dummy-run`` : run configuration parser and datasets generation without benchmarks running.
	* ``--dummy-run``: run configuration parser and dataset generation without benchmarks running.

	* ``report``: create an Excel report based on benchmarks results. Need library `openpyxl`.
	* ``--report``: create an Excel report based on benchmark results. The `openpyxl` library is required.

	* ``output-file``: output file name for result benchmarks. Default is `result.json`
	* ``--output-file``: output file name for the benchmark result. The default name is `result.json`.

	* ``no-intel-optimized`` : using Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). Default running with using Intel(R) Extension for Scikit-learn.
	* ``--no-intel-optimized``: use Scikit-learn without [Intel(R) Extension for Scikit-learn*](#intelr-extension-for-scikit-learn-support). Now available for [scikit-learn benchmarks](https://github.com/IntelPython/scikit-learn_bench/tree/master/sklearn_bench). By default, the runner uses Intel(R) Extension for Scikit-learn.

-The runs of Scikit-learn benchmark use [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) on the CPU by default (use ``no-intel-optimized`` option to run without extention). Some benchmarks have a GPU support:
+When you run scikit-learn benchmarks on CPU, [Intel(R) Extension for Scikit-learn](https://github.com/intel/scikit-learn-intelex) is used by default. Use the ``--no-intel-optimized`` option to run the benchmarks without the extension.
+The following benchmarks have a GPU support:

	A configuration file that contains all these benchmarks can be found [here](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json). You can use this file to run these benchmarks on both CPU and GPU.
	You may use the [configuration file for these benchmarks](https://github.com/IntelPython/scikit-learn_bench/blob/master/configs/skl_xpu_config.json) to run them on both CPU and GPU.

Add GPU benchmarks support to readme #59

Add GPU benchmarks support to readme #59

Conversation

vlad-nazarov commented Mar 30, 2021

michael-smirnov left a comment

Choose a reason for hiding this comment

PetrovKP Mar 30, 2021 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

PetrovKP Mar 30, 2021 •

edited

Loading