Hi, I’m struggling to run jaxlib with cuda support in the new jetpack 6 which has cuda 12.
pip install --upgrade jax==0.4.19 jaxlib==0.4.19+cuda12.cudnn89 -f https://storage.googleapis.com/jax-releases/jax_cuda_releases.html
When I run my python script, which is basically a program that runs whisper-jax I’m getting this message:
CUDA backend failed to initialize: Unable to load cuPTI. Is it installed? (Set TF_CPP_MIN_LOG_LEVEL=0 and rerun for more info.)
and the work is being done with cpu, however I need gpu.
When I run:
nvcc --version
I’m getting:
(.myenv) orin@ubuntu:~/Documents/whisperjax$ nvcc --version
bash: nvcc: command not found
And when tried to install nvidia-cuda-toolkit:
sudo apt install nvidia-cuda-toolkit
this is throwing the next message:
Reading package lists... Done
Building dependency tree... Done
Reading state information... Done
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:
The following packages have unmet dependencies:
nvidia-cuda-toolkit : Depends: nvidia-cuda-dev (= 11.5.1-1ubuntu1) but 6.0-b52 is to be installed
Recommends: nvidia-cuda-toolkit-doc (= 11.5.1-1ubuntu1) but it is not going to be installed
Recommends: nvidia-cuda-gdb (= 11.5.114~11.5.1-1ubuntu1) but it is not going to be installed
Recommends: nsight-compute (= 2021.3.1.4~11.5.1-1ubuntu1)
Recommends: nsight-systems (= 2021.3.3.2~11.5.1-1ubuntu1)
E: Unable to correct problems, you have held broken packages.
After this I have ran (sudo apt-get update, sudo apt-upgrade, sudo apt-get dist-upgrade and also 1.
sudo apt --fix-broken install) and then execute again. However the same message appears.
In my ~/.bashrc and the end I have:
export PATH=/usr/local/cuda-12.2/bin:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-12.2/lib64:$LD_LIBRARY_PATH
I don’t know what to do.
I installed the jetpack 6 successfully through sdk manager by using a virtual machine with ubuntu 20.04
Please help me to fix this problems with nvcc and nvidia-cuda-toolkit installation.
Thanks in advance.