CODEWITHSUNDEEP
tensorflowwindows-subsystem-for-linuxcudnnnuma
nazim elhadi
nazim elhadi

Reputation: 29

Error message in TensorFlow: "could not open file to read NUMA node" and missing directory in /sys/bus/pci/devices

I'm using TensorFlow in my project, and every time I run my code, I get the following error message:

2023-02-23 13:17:55.003041: I tensorflow/compiler/xla/stream_executor/cuda/cuda_gpu_executor.cc:967] could not open file to read NUMA node: /sys/bus/pci/devices/0000:2b:00.0/numa_node Your kernel may have been built without NUMA support.

The message repeats several times in the console, and also mentions that the kernel may not have been built with NUMA support. However, when I check my /sys/bus/pci/devices directory, I cannot find the directory 0000:2b:00.0 mentioned in the error message. Instead, I see the following directories: 74fd:00:00.0 a5cb:00:00.0 c4db:00:00.0 e718:00:00.0 f908:00:00.0

I am using Windows11/WSL2, tensorflow==2.11, python3.9, cudnn==8.1.0.77, nvidia-cudnn==8.2.0.51, cudatoolkit==11.2.2, nvidia-cuda-nvrtc==11.3.58, nvidia-cuda-runtime==11.3.58

Does anyone know why I'm getting this error message and what I can do to resolve it?

Upvotes: 1

Views: 11357

Answers (1)

TF_Renu Patel
TF_Renu Patel

Reputation: 477

As mentioned by @Dr. Snoopy, These are just informational messages which will not interrupt the code execution.

However please verify if you have followed all the steps properly to install tensorflow.

Also please install the specific version of CUDA and cuDNN compatible to Tensorflow 2.11 to have GPU support enabled in your system.

Please refer to this tested build configuration for the same. enter image description here

Upvotes: 0

Related Questions