Error in importing tensorflow in l4t-tensorflow:r35.3.1-tf2.11-py3 container

[sanjay.bharti]

I’m running l4t-tensorflow:r35.3.1-tf2.11-py3 container, inside the container whenever i tried to import tensorflow getting this error:
W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:65] Could not load dynamic library ‘libnvinfer.so.8’; dlerror: /usr/lib/aarch64-linux-gnu/tegra/libnvdla_compiler.so: file too short; LD_LIBRARY_PATH: /usr/lib/aarch64-linux-gnu/tegra:/usr/lib/aarch64-linux-gnu:/usr/local/cuda/lib64
2024-10-03 14:52:34.404992: W tensorflow/compiler/xla/stream_executor/platform/default/dso_loader.cc:65] Could not load dynamic library ‘libnvinfer_plugin.so.8’; dlerror: /usr/lib/aarch64-linux-gnu/tegra/libnvdla_compiler.so: file too short; LD_LIBRARY_PATH: /usr/lib/aarch64-linux-gnu/tegra:/usr/lib/aarch64-linux-gnu:/usr/local/cuda/lib64
2024-10-03 14:52:34.405050: W tensorflow/compiler/tf2tensorrt/utils/py_utils.cc:38] TF-TRT Warning: Cannot dlopen some TensorRT libraries. If you would like to use Nvidia GPU with TensorRT, please make sure the missing libraries mentioned above are installed properly.

[AastaLLL]

Hi,

Do you setup the device with JetPack 6?
Please note that you will need to use the container with the same BSP version.

Thanks.

[sanjay.bharti]

yes, I’ve setup with jetpack 6. How to use container with same BSP version, I’m using latest release of l4t-tensorflow image

[AastaLLL]

Hi,

The BSP in JetPack is r36.

From JetPack 6, please use the TensorFlow container instead of l4t-tensorflow.
For example nvcr.io/nvidia/tensorflow:24.09-tf2-py3-igpu

TensorFlow | NVIDIA NGC

Thanks.