using snpe v2.5.0, Ubuntu 18.04 in WSL2 environment and android ndk r17c.
I am trying to convert the inception v3 .pb model to .dlc using the snpe-tensorflow-to-dlc script (following the tutorial in documentation). it seems the conversion is successful but the .dlc file is not getting written, giving me a segmentation fault at the end...
(myenv) prithviraj@LAPTOP-F4BRR7U4:/mnt/d/internStuff/snpe-2.5.0.4052/models/inception_v3/tensorflow$ snpe-tensorflow-to-dlc -i inception_v3_2016_08_28_frozen.pb --input_dim input "1,299,299,3" --out_node "InceptionV3/Predictions/Reshape_1" -o ../2023-02-23 12:32:07.846322: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'libcudart.so.10.1'; dlerror: libcudart.so.10.1: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /mnt/d/internStuff/snpe-2.5.0.4052/lib/x86_64-linux-clang:/mnt/d/internStuff/snpe-2.5.0.4052/lib/x86_64-linux-clang:2023-02-23 12:32:07.846389: 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.2023-02-23 12:32:59.642489: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations: AVX2 FMATo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.2023-02-23 12:32:59.787157: I tensorflow/core/platform/profile_utils/cpu_utils.cc:104] CPU Frequency: 2112005000 Hz2023-02-23 12:32:59.791177: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x3311810 initialized for platform Host (this does not guarantee that XLA will be used). Devices:2023-02-23 12:32:59.791251: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor device (0): Host, Default Version2023-02-23 12:32:59.821259: W tensorflow/stream_executor/platform/default/dso_loader.cc:59] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /mnt/d/internStuff/snpe-2.5.0.4052/lib/x86_64-linux-clang:/mnt/d/internStuff/snpe-2.5.0.4052/lib/x86_64-linux-clang:2023-02-23 12:32:59.821328: W tensorflow/stream_executor/cuda/cuda_driver.cc:312] failed call to cuInit: UNKNOWN ERROR (303)2023-02-23 12:32:59.821350: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (LAPTOP-F4BRR7U4): /proc/driver/nvidia/version does not exist2023-02-23 12:33:04,061 - 214 - INFO - INFO_ALL_BUILDING_NETWORK:==============================================================Building Network==============================================================2023-02-23 12:33:04,712 - 214 - INFO - Resolving static sub-graphs in network...2023-02-23 12:33:04,892 - 214 - INFO - Resolving static sub-graphs in network, complete.2023-02-23 12:33:05,295 - 214 - INFO - Processed 0 quantization encodings2023-02-23 12:33:05,885 - 214 - INFO - INFO_INITIALIZATION_SUCCESS:2023-02-23 12:33:06,011 - 214 - INFO - INFO_CONVERSION_SUCCESS: Conversion completed successfullySegmentation fault
Dear developer,
WSL is unstable envionment for SNPE converter. Could you please help to conver in ubuntu1804 host or docker?
BR.
Wei