Skip to content

Commit 87e7cf4

Browse files
onecatcnonecatcn
authored
Update install_NGC_PaddlePaddle_ch.rst and add install_NGC_PaddlePaddle_en.rst (#5105)
* Update instalL_NGC_PaddlePaddle_ch.rst replace the saying “NGC PaddlePadddle容器” with “NGC飞桨容器” * Create install_NGC_PaddlePaddle_eg.rst add an english version of the doc on NGC Paddle installation guide * Update index_en.rst add the link to NGC PaddlePaddle installation guide. * Rename install_NGC_PaddlePaddle_eg.rst to install_NGC_PaddlePaddle_en.rst * Update index_en.rst
1 parent 1315a16 commit 87e7cf4

File tree

3 files changed

+127
-6
lines changed

3 files changed

+127
-6
lines changed

‎docs/install/index_en.rst‎

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -214,4 +214,5 @@ The second way to install: compile and install with source code
214214
pip/frompip_en.rst
215215
compile/fromsource_en.rst
216216
install_Kunlun_en.md
217+
install_NGC_PaddlePaddle_en.rst
217218
Tables_en.md

‎docs/install/instalL_NGC_PaddlePaddle_ch.rst‎

Lines changed: 6 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -1,20 +1,20 @@
11
.. _install_NGC_PaddlePaddle_container introduction:
22

33
================================
4-
NGC PaddlePaddle 容器安装指南
4+
NGC飞桨容器安装指南
55
================================
66

77
----------------------
88
整体介绍
99
----------------------
1010

11-
NGC PaddlePaddle 容器针对 NVIDIA GPU 加速进行了优化，并包含一组经过验证的库，可启用和优化NVIDIA GPU 性能。此容器还可能包含对 PaddlePaddle 源代码的修改，以最大限度地提高性能和兼容性。此容器还包含用于加速 ETL (`DALI <https://developer.nvidia.com/dali/>`_, `RAPIDS <https://rapids.ai/>`_),、训练(`cuDNN <https://developer.nvidia.com/cudnn>`_, `NCCL <https://developer.nvidia.com/nccl>`_)和推理(`TensorRT <https://docs.nvidia.com/deeplearning/frameworks/tf-trt-user-guide/index.html>`_)工作负载的软件。
11+
NGC飞桨容器针对 NVIDIA GPU 加速进行了优化，并包含一组经过验证的库，可启用和优化NVIDIA GPU 性能。此容器还可能包含对 PaddlePaddle 源代码的修改，以最大限度地提高性能和兼容性。此容器还包含用于加速 ETL (`DALI <https://developer.nvidia.com/dali/>`_, `RAPIDS <https://rapids.ai/>`_),、训练(`cuDNN <https://developer.nvidia.com/cudnn>`_, `NCCL <https://developer.nvidia.com/nccl>`_)和推理(`TensorRT <https://docs.nvidia.com/deeplearning/frameworks/tf-trt-user-guide/index.html>`_)工作负载的软件。
1212

1313
----------------------
1414
环境准备
1515
----------------------
1616

17-
使用 NGC PaddlePaddle 容器需要主机系统安装以下内容：
17+
使用 NGC飞桨容器需要主机系统安装以下内容：
1818

1919
* `Docker引擎 <https://docs.docker.com/get-docker/>`_
2020

@@ -82,7 +82,7 @@ PaddlePaddle 通过将其作为 Python 模块导入来运行：
8282
NGC容器介绍
8383
----------------------
8484

85-
有关内容的完整列表，请参阅 `NVIDIA PaddlePaddle 容器发行说明 <https://docs.nvidia.com/deeplearning/frameworks/paddle-paddle-release-notes/index.html>`_。
85+
有关内容的完整列表，请参阅 `NGC飞桨容器发行说明 <https://docs.nvidia.com/deeplearning/frameworks/paddle-paddle-release-notes/index.html>`_。
8686
此容器映像包含 NVIDIA 版 PaddlePaddle 的完整源代码，位于 /opt/paddle/paddle。它是作为系统 Python 模块预构建和安装的。
8787
NVIDIA PaddlePaddle 容器针对与 NVIDIA GPU 一起使用进行了优化，并包含以下用于 GPU 加速的软件：
8888

@@ -104,9 +104,9 @@ NVIDIA PaddlePaddle 容器针对与 NVIDIA GPU 一起使用进行了优化，并
104104

105105

106106
--------------------------------------------
107-
NGC PaddlePaddle 容器软件许可协议
107+
NGC飞桨容器软件许可协议
108108
--------------------------------------------
109109

110-
当您下载或使用NGC PaddlePaddle 容器时，即表示您已经同意并接受此 `最终用户许可协议 <https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license>`_ 的条款及其对应约束。
110+
当您下载或使用NGC飞桨容器时，即表示您已经同意并接受此 `最终用户许可协议 <https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license>`_ 的条款及其对应约束。
111111

112112

‎docs/install/install_NGC_PaddlePaddle_en.rst‎

Lines changed: 120 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,120 @@
1+
.. _install_NGC_PaddlePaddle_container introduction:
2+
3+
==============================================
4+
NGC PaddlePaddle Container Installation Guide
5+
==============================================
6+
7+
----------------------
8+
Overview
9+
----------------------
10+
11+
The PaddlePaddle NGC Container is optimized for GPU acceleration, and contains a validated set of libraries that enable and optimize GPU performance. This container may also contain modifications to the PaddlePaddle source code in order to maximize performance and compatibility. This container also contains software for accelerating ETL (`DALI <https://developer.nvidia.com/dali/>`_, `RAPIDS <https://rapids.ai/>`_), Training(`cuDNN <https://developer.nvidia.com/cudnn>`_, `NCCL <https://developer.nvidia.com/nccl>`_), and Inference(`TensorRT <https://docs.nvidia.com/deeplearning/frameworks/tf-trt-user-guide/index.html>`_) workloads。
12+
13+
------------------------------------------
14+
Environmental preparation
15+
------------------------------------------
16+
17+
Using the PaddlePaddle NGC Container requires the host system to have the following installed:
18+
19+
* `Docker Engine <https://docs.docker.com/get-docker/>`_
20+
21+
* `NVIDIA GPU Drivers <https://docs.nvidia.com/datacenter/tesla/tesla-installation-notes/index.html>`_
22+
23+
* `NVIDIA Container Toolkit <https://github.com/NVIDIA/nvidia-docker>`_
24+
25+
For supported versions, see the `Framework Containers Support Matrix <https://docs.nvidia.com/deeplearning/frameworks/support-matrix/index.html>`_ and the `NVIDIA Container Toolkit Documentation <https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/install-guide.html>`_ .
26+
27+
No other installation, compilation, or dependency management is required. It is not necessary to install the NVIDIA CUDA Toolkit.
28+
29+
----------------------
30+
Installation
31+
----------------------
32+
33+
To run a container, issue the appropriate command as explained in the `Running A Container <https://docs.nvidia.com/deeplearning/frameworks/user-guide/index.html#runcont>`_ chapter in the NVIDIA Containers For Deep Learning Frameworks User’s Guide and specify the registry, repository, and tags. For more information about using NGC, refer to the `NGC Container User Guide <https://docs.nvidia.com/ngc/ngc-catalog-user-guide/index.html>`_ .
34+
If you have Docker 19.03 or later, a typical command to launch the container is:
35+
36+
37+
::
38+
39+
docker run --gpus all -it --rm nvcr.io/nvidia/paddlepaddle:22.07-py3
40+
41+
42+
If you have Docker 19.02 or earlier, a typical command to launch the container is:
43+
44+
45+
::
46+
47+
nvidia-docker run -it --rm nvcr.io/nvidia/paddlepaddle:22.07-py3
48+
49+
50+
51+
Where:
52+
53+
* 22.07 is the container version.
54+
55+
PaddlePaddle is run by importing it as a Python module:
56+
57+
58+
::
59+
60+
$ python -c 'import paddle; paddle.utils.run_check()'
61+
Running verify PaddlePaddle program ...
62+
W0516 06:36:54.208734 442 device_context.cc:451] Please NOTE: device: 0, GPU Compute Capability: 8.0, Driver API Version: 11.7, Runtime API Version: 11.7
63+
W0516 06:36:54.212574 442 device_context.cc:469] device: 0, cuDNN Version: 8.4.
64+
PaddlePaddle works well on 1 GPU.
65+
W0516 06:37:12.706600 442 fuse_all_reduce_op_pass.cc:76] Find all_reduce operators: 2. To make the speed faster, some all_reduce ops are fused during training, after fusion, the number of all_reduce ops is 2.
66+
PaddlePaddle works well on 8 GPUs.
67+
PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.
68+
69+
See /workspace/README.md inside the container for information on getting started and customizing your PaddlePaddle image.
70+
71+
You might want to pull in data and model descriptions from locations outside the container for use by PaddlePaddle. To accomplish this, the easiest method is to mount one or more host directories as `Docker bind mounts <https://docs.docker.com/storage/bind-mounts/>`_. For example:
72+
73+
::
74+
75+
docker run --gpus all -it --rm -v local_dir:container_dir nvcr.io/nvidia/paddlepaddle:22.07-py3
76+
77+
78+
Note: In order to share data between ranks, NCCL may require shared system memory for IPC and pinned (page-locked) system memory resources. The operating system's limits on these resources may need to be increased accordingly. Refer to your system's documentation for details. In particular, Docker containers default to limited shared and pinned memory resources. When using NCCL inside a container, it is recommended that you increase these resources by issuing:
79+
80+
81+
::
82+
83+
--shm-size=1g --ulimit memlock=-1
84+
85+
in the docker run command.
86+
87+
88+
------------------------------------------
89+
Introduction to NGC Container
90+
------------------------------------------
91+
92+
For the full list of contents, see the `NGC PaddlePaddle Container Release Notes <https://docs.nvidia.com/deeplearning/frameworks/paddle-paddle-release-notes/index.html>`_.
93+
94+
This container image contains the complete source of the NVIDIA version of PaddlePaddle in /opt/paddle/paddle. It is prebuilt and installed as a system Python module. Visit paddlepaddle.org.cn to learn more about PaddlePaddle.
95+
96+
The NVIDIA PaddlePaddle Container is optimized for use with NVIDIA GPUs, and contains the following software for GPU acceleration:
97+
98+
99+
* `CUDA <https://developer.nvidia.com/cuda-toolkit>`_
100+
101+
* `cuBLAS <https://developer.nvidia.com/cublas>`_
102+
103+
* `NVIDIA cuDNN <https://developer.nvidia.com/cudnn>`_
104+
105+
* `NVIDIA NCCL <https://developer.nvidia.com/nccl>`_ (optimized for `NVLink <http://www.nvidia.com/object/nvlink.html>`_ )
106+
107+
* `NVIDIA Data Loading Library (DALI) <https://developer.nvidia.com/dali>`_
108+
109+
* `TensorRT <https://developer.nvidia.com/tensorrt>`__
110+
111+
* `PaddlePaddle with TensorRT (Paddle-TRT) <https://github.com/PaddlePaddle/Paddle-Inference-Demo/blob/master/docs/optimize/paddle_trt_en.rst>`_
112+
113+
The software stack in this container has been validated for compatibility, and does not require any additional installation or compilation from the end user. This container can help accelerate your deep learning workflow from end to end.
114+
115+
116+
--------------------------------------------
117+
License
118+
--------------------------------------------
119+
120+
By pulling and using the container, you accept the terms and conditions of this `End User License Agreement <https://developer.nvidia.com/ngc/nvidia-deep-learning-container-license>`_ .

0 commit comments

Comments
 (0)