That’s a odd error. If you just want to test inferences, can you use pytorch for CPU extraction instead? If the text extraction is good enough, then you can debug building cuda later.
1.13 <= PyTorch or 2.11 <= Tensorflow <2.16. On lower Tensorflow versions the code will only run inference on a GPU. In general, if you want to train or fine-tune models, a GPU is required.
With respect to the deep learning framework, you must decide between Tensorflow and PyTorch .
We use Pillow or OpenCV for image processing tasks. PIL is more lightweight, easier to install and the default choice. OpenCV is faster when loading images and can be beneficial especially when training. If you want to use OpenCV, please install this framework separately and set the environment variable USE_OPENCV=True.
You can run on PyTorch with a CPU only. For Tensorflow a GPU is required.
On the topic of cuda, can you check that the nvidia proprietary drivers are installed? OP had it
Steps I took
# first fully update 22.04 LTS
apt update && apt upgrade -y
# reboot ... you probably got a newer kernel...
# ensure remote access
Since we are updating the video driver, and it is likely you don't have more than one gpu in the system, ensure you can ```ssh``` into the system from another system. This is useful for both setup and troubleshooting, Should Something Go Wrong.
# nvidia part 1
We need the nvidia GPU proprietary driver first. If the only GPU in the system is nvidia and you're using the nouveau driver, it must be blacklisted first. Before you reboot, install the nvidia drivers. Then reboot.
```lsmod``` and check the output to confirm the ```nvidia``` module is loaded; check ```dmesg``` to be sure you do NOT see messages like:
[ 1044.501389] NVRM: The NVIDIA probe routine was not called for 1 device(s).
... this message indicates "something else" has claimed your nvidia card (most likely nouveau).
Once nvidia is loaded and ```dmesg``` is free of errors that might indicate the nvidia driver
# nvidia part 2
wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-keyring_1.1-1_all.deb
sudo apt update
The package not found error is related to not having the cuda repo
Also if you want to post the text as a code block ``` \n code output here \n ``` to make it easier you can use markdown formatting on discourse
1 Like