python – Native Tensorflow GPU on Fedora linux

admin

Please help me!
I need to install tensorflow using GPU computing.
Here is my nvidia-smi:
nvidia-smi

Which version of Python should I use?
conda create --name gpu python=????

conda install -c conda-forge cudatoolkit=???? cudnn=????

In general, how do I do this correct?

I’ve used this before:

conda create --name gpu python=3.9
conda install -c conda-forge cudatoolkit=11.2 cudnn=8.1.0
pip install tensorflow==2.8.0

However, it didn’t help me to use the GPU:

(gpu) ferz@ferz:~$ python3 -c "import tensorflow as tf; print(tf.config.list_physical_devices('GPU'))"
2024-05-26 16:59:34.944925: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2024-05-26 16:59:34.944947: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.
2024-05-26 16:59:36.697217: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:936] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2024-05-26 16:59:36.697421: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697486: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcublas.so.11'; dlerror: libcublas.so.11: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697540: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcublasLt.so.11'; dlerror: libcublasLt.so.11: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697591: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcufft.so.10'; dlerror: libcufft.so.10: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697641: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcurand.so.10'; dlerror: libcurand.so.10: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697691: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcusolver.so.11'; dlerror: libcusolver.so.11: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697740: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcusparse.so.11'; dlerror: libcusparse.so.11: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697790: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcudnn.so.8'; dlerror: libcudnn.so.8: cannot open shared object file: No such file or directory
2024-05-26 16:59:36.697801: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1850] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.
Skipping registering GPU devices...
[]

Read more here: Source link