Hello Guys,
Im having to run older ML code that is written for Cuda 10. I checked the compatibility list and i need to install NV drivers 410 to get Cuda 10. I tried really hard to install it and i cant get it done. I also need to run newer code too. So i would need two Operating systems one to run the newest drivers and one 410 and Cuda 10. Im trying to solve this with Containers.
I installed Silverblue and i created two containers. I succesfully installed the newest drivers in one and im now trying to install drivers 410 and Cuda 10 on the second but i cant do it.
Os in that container is Fedora 29.
Any advice how i can install 410 and Cuda 10 on fedora 29?
I have downloaded cuda 10 from here:
https://developer.nvidia.com/cuda-10.0-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Fedora&target_version=27&target_type=rpmlocal
and i tried to install it with
sudo yum localinstall cuda-repo-fedora27-10-0-local-10.0.130-410.48-1.0-1.x86_64.rpm
When i check with yum list | grep cuda i find stuff but i cant use
nvcc --version
to check the version. Perhaps because i havent installed the Nvidia driver yet.
Here is the driver installation attempt.
I found the right driver but i cant execute the run inside the container (fedora).
ERROR: Unable to find the module utility `modprobe`; please make sure you have the package 'module-init-tools' or 'kmod' installed. If you do have 'module-init-tools' or 'kmod' installed, then please check that `modprobe` is in your
PATH.
Any adive? is there another way besides the run file to install 410.57 from an RPM?
Please help. Im trying to solve this since a week and i cant manage to do it. Im so behind my project in ML. Please someone any idea?
EDIT:
I installed a package named:
module-init-tools
with
sudo yum install module-init-tools
And now im getting another error:
ERROR: An NVIDIA kernel module 'nvidia-uvm' appears to already be loaded in your kernel. This may be because it
is in use (for example, by an X server, a CUDA program, or the NVIDIA Persistence Daemon), but this may
also happen if your kernel was configured without support for module unloading. Please be sure to exit any
programs that may be using the GPU(s) before attempting to upgrade your driver. If no GPU-based programs
are running, you know that your kernel supports module unloading, and you still receive this message, then
an error may have occured that has corrupted an NVIDIA kernel module's usage count, for which the simplest
remedy is to reboot your computer.