AGX Xavier ImportError: libcublas.so.10: cannot open shared object file: No such file or directory

Hello, I’m currently working on a xavier AGX with Jetpack version 5.0.2, but because of lack of hard disk space I can’t install l4t-ml:r35.1.0-py3 image on the xavier, so I’m using l4t-ml:r32.7.1-py3 (for Jetpack Version 4.6.1).

When running the container with sudo docker run -it --rm --runtime nvidia --network host nvcr.io/nvidia/l4t-ml:r32.7.1-py3
and trying to import tensorflow inside of python
/# python3
>>> import tensorflow
I get the following error:

2022-10-07 08:00:35.878373: 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.
2022-10-07 08:00:35.878827: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'libcudart.so.10.2'; dlerror: libcudart.so.10.2: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda/lib64:/usr/local/cuda-10.2/targets/aarch64-linux/lib:
2022-10-07 08:00:35.878902: 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.
Segmentation fault (core dumped)

My guess is that due to version mismatch this error occurs, am I right in my assumption?

Hi,

Please noted that there are some dependencies between OS and libraries.
You will need to use the container with r35.1 tag for JetPack 5.0.2.

Thanks.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.