How to use PyTorch with Container Station?
What is PyTorch?
It’s a Python-based scientific computing package targeted at two sets of audiences:
Installation Instructions
Recommended version
QTS 4.3.5/4.3.6 and Nvidia Driver qpkg v1.3.5
Tag
|
Pull command
|
Description
|
pytorch/pytorch:0.4.1-cuda9-cudnn7-devel
|
docker pull pytorch/pytorch:0.4.1-cuda9-cudnn7-devel
|
|
QTS 4.4.x and Nvidia Driver qpkg v2.0.0
Tag
|
Pull command
|
Description
|
pytorch/pytorch:1.0.1-cuda10.0-cudnn7-devel
|
docker pull pytorch/pytorch:1.0.1-cuda10.0-cudnn7-devel
|
|
Before running the PyTorch container, use docker pull command or click the Pull Button to ensure the desired image is installed. Once the pull is complete, you can go to next step.
Go to Control Panel -> Hardware -> Graphics Card. Assign the GPU to the Container Station for Resource Use.
Go to Container Station > Click "Create". Search "PyTorch " and click "Install/Create".
Using Command to mount NVIDIA GPU
Using SSH to NAS and then can follow the below steps:
Add below commands when running Docker Run:
--device /dev/nvidia0:/dev/nvidia0 \
--device /dev/nvidiactl:/dev/nvidiactl \
--device /dev/nvidia-uvm:/dev/nvidia-uvm \
-v `/sbin/getcfg NVIDIA_GPU_DRV Install_Path -f /etc/config/qpkg.conf -d None`/usr/:/usr/local/nvidia
related information:
Comments
Post a Comment