Hendrik/xgboost
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Small updates to GPU documentation (#5483)

Browse Source
This commit is contained in:
Rory Mitchell 2020-04-05 08:02:27 +12:00 committed by GitHub
parent a9313802ea
commit 15800107ad
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 14 additions and 21 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
doc/gpu/index.rst
View File

@ -26,7 +26,7 @@ Algorithms
+-----------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------+
| tree_method | Description |
+=======================+=======================================================================================================================================================================+
| gpu_hist | Equivalent to the XGBoost fast histogram algorithm. Much faster and uses considerably less memory. NOTE: Will run very slowly on GPUs older than Pascal architecture. |
| gpu_hist | Equivalent to the XGBoost fast histogram algorithm. Much faster and uses considerably less memory. NOTE: May run very slowly on GPUs older than Pascal architecture. |
+-----------------------+-----------------------------------------------------------------------------------------------------------------------------------------------------------------------+
Supported parameters
@ -50,8 +50,6 @@ Supported parameters
+--------------------------------+--------------+
| ``gpu_id`` | |tick| |
+--------------------------------+--------------+
| ``n_gpus`` (deprecated) | |tick| |
+--------------------------------+--------------+
| ``predictor`` | |tick| |
+--------------------------------+--------------+
| ``grow_policy`` | |tick| |
@ -85,10 +83,6 @@ The GPU algorithms currently work with CLI, Python and R packages. See :doc:`/bu
XGBRegressor(tree_method='gpu_hist', gpu_id=0)
Single Node Multi-GPU
=====================
.. note:: Single node multi-GPU training with `n_gpus` parameter is deprecated after 0.90. Please use distributed GPU training with one process per GPU.
Multi-node Multi-GPU Training
=============================
XGBoost supports fully distributed GPU training using `Dask <https://dask.org/>`_. For
@ -128,11 +122,11 @@ Most of the objective functions implemented in XGBoost can be run on GPU. Follo
+--------------------+-------------+
| survival:cox | |cross| |
+--------------------+-------------+
| rank:pairwise | |cross| |
| rank:pairwise | |tick| |
+--------------------+-------------+
| rank:ndcg | |cross| |
| rank:ndcg | |tick| |
+--------------------+-------------+
| rank:map | |cross| |
| rank:map | |tick| |
+--------------------+-------------+
Objective will run on GPU if GPU updater (``gpu_hist``), otherwise they will run on CPU by
@ -160,13 +154,13 @@ Following table shows current support status for evaluation metrics on the GPU.
+-----------------+-------------+
| mlogloss | |tick| |
+-----------------+-------------+
| auc | |cross| |
| auc | |tick| |
+-----------------+-------------+
| aucpr | |cross| |
+-----------------+-------------+
| ndcg | |cross| |
| ndcg | |tick| |
+-----------------+-------------+
| map | |cross| |
| map | |tick| |
+-----------------+-------------+
| poisson-nloglik | |tick| |
+-----------------+-------------+
@ -188,21 +182,18 @@ You can run benchmarks on synthetic data for binary classification:
.. code-block:: bash
python tests/benchmark/benchmark.py
python tests/benchmark/benchmark_tree.py --tree_method=gpu_hist
python tests/benchmark/benchmark_tree.py --tree_method=hist
Training time time on 1,000,000 rows x 50 columns with 500 boosting iterations and 0.25/0.75 test/train split on i7-6700K CPU @ 4.00GHz and Pascal Titan X yields the following results:
Training time on 1,000,000 rows x 50 columns of random data with 500 boosting iterations and 0.25/0.75 test/train split with AMD Ryzen 7 2700 8 core @3.20GHz and Nvidia 1080ti yields the following results:
+--------------+----------+
| tree_method | Time (s) |
+==============+==========+
| gpu_hist | 13.87 |
| gpu_hist | 12.57 |
+--------------+----------+
| hist | 63.55 |
| hist | 36.01 |
+--------------+----------+
| exact | 1082.20 |
+--------------+----------+
See `GPU Accelerated XGBoost <https://xgboost.ai/2016/12/14/GPU-accelerated-xgboost.html>`_ and `Updates to the XGBoost GPU algorithms <https://xgboost.ai/2018/07/04/gpu-xgboost-update.html>`_ for additional performance benchmarks of the ``gpu_hist`` tree method.
Memory usage
============
@ -241,8 +232,10 @@ Many thanks to the following contributors (alphabetical order):
* Jonathan C. McKinney
* Matthew Jones
* Philip Cho
* Rong Ou
* Rory Mitchell
* Shankara Rao Thejaswi Nanditale
* Sriram Chandramouli
* Vinay Deshpande
Please report bugs to the XGBoost issues list: https://github.com/dmlc/xgboost/issues. For general questions please visit our user form: https://discuss.xgboost.ai/.