# Failed to load the native TensorFlow runtime. **URL:** \ **Category:** TAO Toolkit\ **Created:** [November 18, 2019, 12:05pm UTC](https://forums.developer.nvidia.com/t/failed-to-load-the-native-tensorflow-runtime/107608 "2019-11-18T12:05:33Z")\ **Posts on this page:** 4\ **Page:** 1
**Author:** ![NitinRai](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/nitinrai/32/15561_2.png) [@NitinRai](https://forums.developer.nvidia.com/u/NitinRai)\ **Post date:** [November 18, 2019, 12:05pm UTC](https://forums.developer.nvidia.com/t/failed-to-load-the-native-tensorflow-runtime/107608/1 "2019-11-18T12:05:33Z")
Trying to convert kitti dataset to tfrecords. command used: ```auto !tlt-dataset-convert -d spec_files/convert.txt -o /tf_records ``` or ```auto !tlt-dataset-convert -d /workspace/spec_files/convert.txt -o /workspace/tf_records/ ``` and many other path combinations. The spec\_file being used: ```auto kitti_config { root_directory_path: "/workspace/dataset/KITTI_original/training" image_dir_name: "image_2" label_dir_name: "label_2" image_extension: ".png" partition_mode: "random" num_partitions:2 val_split: 15 num_shards: 10 } image_directory_path: "/workspace/dataset/KITTI_original/training" ``` Error: ```auto Traceback (most recent call last): File "/usr/local/bin/tlt-dataset-convert", line 6, in from iva.detectnet_v2.scripts.dataset_convert import main File "./detectnet_v2/scripts/dataset_convert.py", line 14, in File "./detectnet_v2/dataio/build_converter.py", line 13, in File "./detectnet_v2/dataio/kitti_converter_lib.py", line 16, in File "/usr/local/lib/python2.7/dist-packages/tensorflow/ __init__.py", line 24, in from tensorflow.python import pywrap_tensorflow # pylint: disable=unused-import File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ __init__.py", line 49, in from tensorflow.python import pywrap_tensorflow File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 74, in raise ImportError(msg) ImportError: Traceback (most recent call last): File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in from tensorflow.python.pywrap_tensorflow_internal import * File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in _pywrap_tensorflow_internal = swig_import_helper() File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, description) ImportError: libcuda.so.1: cannot open shared object file: No such file or directory Failed to load the native TensorFlow runtime. See https://www.tensorflow.org/install/errors for some common reasons and solutions. Include the entire stack trace above this error message when asking for help. ``` trying to follow this notebook, [https://github.com/NVIDIA-AI-IOT/tlt-iva-examples/blob/master/TLT\_notebook.ipynb](https://github.com/NVIDIA-AI-IOT/tlt-iva-examples/blob/master/TLT_notebook.ipynb) kindly look into it. ---
**Author:** ![Ravik](https://developer.download.nvidia.com/images/forums/profile-default-devtalk-84.png) [@Ravik](https://forums.developer.nvidia.com/u/Ravik)\ **Post date:** [November 18, 2019, 12:23pm UTC](https://forums.developer.nvidia.com/t/failed-to-load-the-native-tensorflow-runtime/107608/2 "2019-11-18T12:23:42Z")
Environment details DGX version ```auto grep VERSION /etc/dgx-release DGX_SWBUILD_VERSION="3.1.4" DGX_OTA_VERSION="3.1.7" DGX_OTA_VERSION="3.1.8" ``` ```auto PATH /opt/nvidia/tools:/opt/ngccli:/usr/local/bin:/usr/local/nvidia/bin:/usr/local/cuda/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin CUDA_VERSION 10.0.130 CUDA_PKG_VERSION 10-0=10.0.130-1 LD_LIBRARY_PATH /usr/local/lib:/usr/local/nvidia/lib:/usr/local/nvidia/lib64 NVIDIA_VISIBLE_DEVICES all NVIDIA_DRIVER_CAPABILITIES compute,utility NVIDIA_REQUIRE_CUDA cuda>=10.0 brand=tesla,driver>=384,driver<385 brand=tesla,driver>=410,driver<411 NCCL_VERSION 2.4.2 LIBRARY_PATH /usr/local/cuda/lib64/stubs MOFED_VERSION 3.4-1.0.0.0 TRT_LIB_PATH /usr/lib/x86_64-linux-gnu TRT_INCLUDE_PATH /usr/include/x86_64-linux-gnu PYTHONPATH /workspace/examples:/opt/nvidia/third_party/nvml:/opt/nvidia: ``` ---
**Author:** ![NitinRai](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/nitinrai/32/15561_2.png) [@NitinRai](https://forums.developer.nvidia.com/u/NitinRai)\ **Post date:** [November 18, 2019, 4:11pm UTC](https://forums.developer.nvidia.com/t/failed-to-load-the-native-tensorflow-runtime/107608/3 "2019-11-18T16:11:04Z")
I found the solution here! sorry for raising new topic. [https://devtalk.nvidia.com/default/topic/1065003/transfer-learning-toolkit/conversion-of-kitti-to-tfrecords-failing-due-to-importerror-usr-lib-x86\_64-linux-gnu-libcuda-so-1-file-too-short/post/5392892/#5392892](https://devtalk.nvidia.com/default/topic/1065003/transfer-learning-toolkit/conversion-of-kitti-to-tfrecords-failing-due-to-importerror-usr-lib-x86_64-linux-gnu-libcuda-so-1-file-too-short/post/5392892/#5392892) ---
**Author:** ![TomNVIDIA](https://sea2.discourse-cdn.com/nvidia/user_avatar/forums.developer.nvidia.com/tomnvidia/32/14181_2.png) [@TomNVIDIA](https://forums.developer.nvidia.com/u/TomNVIDIA)\ **Post date:** [October 12, 2021, 4:24pm UTC](https://forums.developer.nvidia.com/t/failed-to-load-the-native-tensorflow-runtime/107608/4 "2021-10-12T16:24:52Z")
Closed on October 12, 2021, 4:24pm UTC