Openvino 在 Ubuntu Docker 中未检测到 Intel GPU
我对 Openvino 和 Intel GPU 世界还很陌生,我已经提取了 openvino/openvino_tensorflow_ubuntu18_runtime:2.0.0 docker 映像,当我运行 OpenVINO_TensorFlow_classification_example.ipynb 时,它在列出后端时只看到“CPU”。
但是,当在 docker 容器内运行 lspci -nn -s 0:002.0
时,它会输出:
00:02.0 VGA compatible controller [0300]: Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller [8086:0412] (rev 06)
我尝试运行一个更简单的 python 脚本来仅打印 openvino 后端,但我始终只获取 CPU 后端。
我在这里缺少什么?我的 iGPU 与 OpenVino 不兼容吗?我是否缺少某些驱动程序或某种配置?
我的本机操作系统是 Debian11。
编辑:
事实上,我按照 Peh_Intel 的建议,使用图像文档中的 iGPU 命令运行了 docker 容器。但是,我始终只获得 CPU 后端。下面是终端和 jupyter 笔记本输出的屏幕截图。
I am pretty new to this Openvino and Intel GPU world, I have pulled openvino/openvino_tensorflow_ubuntu18_runtime:2.0.0 docker image and when I run the OpenVINO_TensorFlow_classification_example.ipynb it only sees "CPU" when listing backends.
However, when running lspci -nn -s 0:002.0
inside the docker container it outputs:
00:02.0 VGA compatible controller [0300]: Intel Corporation Xeon E3-1200 v3/4th Gen Core Processor Integrated Graphics Controller [8086:0412] (rev 06)
I have tried to run a simpler python script to only print the openvino backends but I keep getting only the CPU backend.
What am I missing here? Is my iGPU not compatible with OpenVino? Am I missing some driver or some kind of configuration?
My native OS is Debian11.
EDIT:
Indeed I ran the docker container with the iGPU command in the image documentation as Peh_Intel suggests. However, I keep getting only the CPU backend. There is an screenshot of terminal and jupyter notebook outputs below.
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(1)
GPU默认情况下在容器中不可用。将其连接到容器上。
使用IGPU访问启动Jupyter服务器:
GPU is not available in the container by default. Attach it to the container.
Launch the Jupyter server with iGPU access: