blocks|key|787879|text|您必须为您的Linux分发安装NVIDIA驱动程序和Docker+19.03。请注意，您不需要在主机上安装CUDA工具包，但需要安装驱动程序|type|unstyled|depth|inlineStyleRanges|entityRanges|data|787880|随着Docker+19.03的发布，nvidia-docker2包的使用被弃用，因为NVIDIA+GPU现在作为Docker运行时中的设备提供本地支持。|787881|如果您使用的是nvidia-docker2包，请查看“Upgrading+with+nvidia-docker2”中的说明。|offset|length|787882|对于首次使用Docker+19.03和GPU的用户，请继续阅读下面的入门说明。Usage|787883|#+Add+the+package+repositories
distribution=$(.+/etc/os-release;echo+$ID$VERSION_ID)
curl+-s+-L+https://nvidia.github.io/nvidia-docker/gpgkey+%7C+sudo+apt-key+add+-
curl+-s+-L+https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list+%7C+sudo+tee+/etc/apt/sources.list.d/nvidia-docker.list

sudo+apt-get+update+&&+sudo+apt-get+install+-y+nvidia-container-toolkit
sudo+systemctl+restart+docker|code-block|syntax|javascript|787884|如需更多信息，请访问：https://github.com/NVIDIA/nvidia-docker。|787885|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/NVIDIA/nvidia-docker/tree/master#upgrading-with-nvidia-docker2-deprecated|1|2|https://github.com/NVIDIA/nvidia-docker^0|0|0|R|T|0|0|13|5|1|0|0|B|13|2|0^^$0|@$1|2|3|4|5|6|7|11|8|@]|9|@]|A|$]]|$1|B|3|C|5|6|7|12|8|@]|9|@]|A|$]]|$1|D|3|E|5|6|7|13|8|@]|9|@$F|14|G|15|1|16]]|A|$]]|$1|H|3|I|5|6|7|17|8|@]|9|@$F|18|G|19|1|1A]]|A|$]]|$1|J|3|K|5|L|7|1B|8|@]|9|@]|A|$M|N]]|$1|O|3|P|5|6|7|1C|8|@]|9|@$F|1D|G|1E|1|1F]]|A|$]]|$1|Q|3|-4|5|6|7|1G|8|@]|9|@]|A|$]]]|R|$S|$5|T|U|V|A|$W|X]]|Y|$5|T|U|V|A|$W|X]]|Z|$5|T|U|V|A|$W|10]]]]

You have to install the NVIDIA driver and Docker 19.03 for your Linux distribution Note that you do not need to install the CUDA toolkit on the host, but the driver needs to be installed

With the release of Docker 19.03, usage of nvidia-docker2 packages are deprecated since NVIDIA GPUs are now natively supported as devices in the Docker runtime.

If you are using the nvidia-docker2 packages, review the instructions in the “<a href="https://github.com/NVIDIA/nvidia-docker/tree/master#upgrading-with-nvidia-docker2-deprecated" rel="nofollow noreferrer">Upgrading with nvidia-docker2</a>”.

For first-time users of Docker 19.03 and GPUs, continue with the instructions for getting started below. <a href="https://github.com/NVIDIA/nvidia-docker/tree/master#upgrading-with-nvidia-docker2-deprecated" rel="nofollow noreferrer">Usage</a>

<pre><code># Add the package repositories
distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list

sudo apt-get update &amp;&amp; sudo apt-get install -y nvidia-container-toolkit
sudo systemctl restart docker
</code></pre>

for more information please visit : <a href="https://github.com/NVIDIA/nvidia-docker" rel="nofollow noreferrer">https://github.com/NVIDIA/nvidia-docker</a>

blocks|key|527513|text|我找到的一个“解决方案”是首先运行一个安装了主机nvidia驱动程序的基础镜像|type|unstyled|depth|inlineStyleRanges|entityRanges|data|527514|docker+run+-it+--rm+--gpus+ubuntu|code-block|syntax|javascript|527515|然后在容器中手动构建我的应用程序，并提交生成的映像。这并不理想，最好在构建阶段访问nvidia-smi。|527516|entityMap^0|0|0|0^^$0|@$1|2|3|4|5|6|7|K|8|@]|9|@]|A|$]]|$1|B|3|C|5|D|7|L|8|@]|9|@]|A|$E|F]]|$1|G|3|H|5|6|7|M|8|@]|9|@]|A|$]]|$1|I|3|-4|5|6|7|N|8|@]|9|@]|A|$]]]|J|$]]

A "solution" I found is to first run a base image with the host nvidia drivers mounted on it

<pre><code>docker run -it --rm --gpus ubuntu
</code></pre>

And then build my app within the container manually and commit the resulting image.
This is not ideal and it would be best to have access to nvidia-smi during the build phase.

blocks|key|527538|text|你需要像docs中解释的那样使用nvidia-container-runtime：“这也是在docker构建过程中访问GPU的唯一方法”。|type|unstyled|depth|inlineStyleRanges|entityRanges|offset|length|data|527539|Ubuntu的步骤：|527540|容器安装nvidia-|527541|-runtime：|ordered-list-item|527542|527543|sudo+apt-get+install+nvidia-container-runtime|style|CODE|527544|Edit/create+/etc/docker/daemon.json的内容：|unordered-list-item|527545|527546|{
++++"runtimes":+{
++++++++"nvidia":+{
++++++++++++"path":+"/usr/bin/nvidia-container-runtime",
++++++++++++"runtimeArgs":+[]
+++++++++}+
++++},
++++"default-runtime":+"nvidia"+
}|code-block|syntax|javascript|527547|527548|重新启动docker守护程序：|527549|527550|sudo+systemctl+restart+docker|527551|Build你的镜像(现在GPU在构建过程中可用)：|527552|527553|docker+build+-t+my_image_name:latest+.|527554|entityMap|0|LINK|mutability|MUTABLE|url|https://github.com/NVIDIA/nvidia-docker/wiki/Advanced-topics#default-runtime|1|https://github.com/nvidia/nvidia-container-runtime#docker-engine-setup^0|4|4|0|G|O|1|0|0|0|0|0|0|19|0|0|0|0|0|0|0|0|T|0|0|0|0|12|0^^$0|@$1|2|3|4|5|6|7|1J|8|@]|9|@$A|1K|B|1L|1|1M]|$A|1N|B|1O|1|1P]]|C|$]]|$1|D|3|E|5|6|7|1Q|8|@]|9|@]|C|$]]|$1|F|3|G|5|6|7|1R|8|@]|9|@]|C|$]]|$1|H|3|I|5|J|7|1S|8|@]|9|@]|C|$]]|$1|K|3|-4|5|6|7|1T|8|@]|9|@]|C|$]]|$1|L|3|M|5|6|7|1U|8|@$A|1V|B|1W|N|O]]|9|@]|C|$]]|$1|P|3|Q|5|R|7|1X|8|@]|9|@]|C|$]]|$1|S|3|-4|5|6|7|1Y|8|@]|9|@]|C|$]]|$1|T|3|U|5|V|7|1Z|8|@]|9|@]|C|$W|X]]|$1|Y|3|-4|5|6|7|20|8|@]|9|@]|C|$]]|$1|Z|3|10|5|J|7|21|8|@]|9|@]|C|$]]|$1|11|3|-4|5|6|7|22|8|@]|9|@]|C|$]]|$1|12|3|13|5|6|7|23|8|@$A|24|B|25|N|O]]|9|@]|C|$]]|$1|14|3|15|5|R|7|26|8|@]|9|@]|C|$]]|$1|16|3|-4|5|6|7|27|8|@]|9|@]|C|$]]|$1|17|3|18|5|6|7|28|8|@$A|29|B|2A|N|O]]|9|@]|C|$]]|$1|19|3|-4|5|6|7|2B|8|@]|9|@]|C|$]]]|1A|$1B|$5|1C|1D|1E|C|$1F|1G]]|1H|$5|1C|1D|1E|C|$1F|1I]]]]

You need use <a href="https://github.com/nvidia/nvidia-container-runtime#docker-engine-setup" rel="noreferrer">nvidia-container-runtime</a> as explained in <a href="https://github.com/NVIDIA/nvidia-docker/wiki/Advanced-topics#default-runtime" rel="noreferrer">docs</a>: "It is also the only way to have GPU access during docker build".

Steps for Ubuntu:

<ol>
<li>Install nvidia-container-runtime:

<code>sudo apt-get install nvidia-container-runtime</code></li>
<li>Edit/create the /etc/docker/daemon.json with content:</li>
</ol>

<pre><code>{
 "runtimes": {
 "nvidia": {
 "path": "/usr/bin/nvidia-container-runtime",
 "runtimeArgs": []
 } 
 },
 "default-runtime": "nvidia" 
}
</code></pre>

<ol start="3">
<li>Restart docker daemon:

<code>sudo systemctl restart docker</code></li>
<li>Build your image (now GPU available during build):

<code>docker build -t my_image_name:latest .</code></li>
</ol>

I have a GPU application that does unit-testing during the image building stage.
With Docker 19.03, one can specify nvidia runtime with <code>docker run --gpus all</code> but I also need access to the gpus for <code>docker build</code> because I do unit-testing. How can I achieve this goal?

For older version of docker that use nvidia-docker2 it was not possible to specifiy runtime during build stage, BUT you can set the default runtime to be nvidia, and docker build works fine that way. Can I do that in Docker 19.03 that doesn't need nvidia-docker anymore? If so, how?

docker build with nvidia runtime

翻译质量差，导致语言生硬或混乱。

没有提供实际的解决方法或示例。

解答不清晰，无法理解或解决问题。

页面排版不美观，阅读体验差。

文章

问答

视频

学习中心

腾讯云实验室

直播

竞赛

腾讯云代码分析专区

腾讯iOA零信任安全管理系统专区

腾讯云架构师技术同盟交流圈

腾讯云数据库专区

腾讯云顾问专区

腾讯云原生专区

腾讯混元专区

腾讯云TCE专区

腾讯云Lighthouse专区

腾讯云HAI专区

腾讯云Edgeone专区

腾讯云存储专区

腾讯云智能专区

腾讯轻联专区 

腾讯云开发专区

TAPD专区

腾讯轻量云游戏服专区

腾讯云最具价值专家

腾讯云架构师技术同盟

腾讯云创作之星

腾讯云开发者先锋

腾讯云代码助手

云原生构建

TAPD 敏捷项目管理

Cloud Studio

SDK中心

API中心

命令行工具

涵盖代码开发、场景应用、自动测试全流程，助你从零构建专属AI助手

一站式MCP教程库，解锁AI应用新玩法

我有一个在映像构建阶段执行单元测试的GPU应用程序。使用Docker19.03，可以使用docker run --gpus all指定nvidia运行时，但我还需要访问docker build的gpus，因为我要做单元测试。我如何才能实现这个目标？对于使用nvidia-docker2的docker的旧版本，无法在构建阶...

问使用nvidia运行时构建docker
EN

Stack Overflow用户

社区

活动

圈层

关于

腾讯云开发者

热门产品

热门推荐

更多推荐

问使用nvidia运行时构建dockerEN