make
../Makefile.config:6: CUDA_INSTALL_DIR variable is not specified, using /usr/local/cuda- by default, use CUDA_INSTALL_DIR=<cuda_directory> to change.
../Makefile.config:9: CUDNN_INSTALL_DIR variable is not specified, using by default, use CUDNN_INSTALL_DIR=<cudnn_directory> to change.
Compiling: sampleMNIST.cpp
sampleMNIST.cpp:9:30: fatal error: cuda_runtime_api.h: File o directory non esistente
compilation terminated.
../Makefile.config:176: set di istruzioni per l'obiettivo "../../bin/dchobj/sampleMNIST.o" non riuscito
make: *** [../../bin/dchobj/sampleMNIST.o] Errore 1
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ tensorrt
RuntimeError: module compiled against API version 0xb but this version of numpy is 0xa
Traceback (most recent call last):
File "/usr/local/bin/tensorrt", line 7, in <module>
from tensorrt.__main__ import main
[b] File "/usr/local/lib/python3.5/dist-packages/tensorrt/__init__.py", line 28, in <module>
from tensorrt import infer, parsers, utils, lite, plugins
File "/usr/local/lib/python3.5/dist-packages/tensorrt/infer/__init__.py", line 5, in <module>
from ._infer_enums import *
File "/usr/local/lib/python3.5/dist-packages/tensorrt/infer/_infer_enums.py", line 5, in <module>
from tensorrt.infer import _nv_infer_bindings as nvinfer[/b]
ImportError: numpy.core.multiarray failed to import
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ which tensorrt
/usr/local/bin/tensorrt
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2017 NVIDIA Corporation
Built on Fri_Sep__1_21:08:03_CDT_2017
Cuda compilation tools, release 9.0, V9.0.176
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ which cu
cudafe cuda-gdbserver cuda-memcheck cupsaddsmb cupsctl cupsenable cupsreject curl
cudafe++ cudaHashcat64.bin cuobjdump cups-browsed cupsd cupsfilter cupstestdsc cut
cuda-gdb cuda-install-samples-9.0.sh cupsaccept cups-calibrate cupsdisable cups-genppdupdate cupstestppd
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ which cuda
cudafe cudafe++ cuda-gdb cuda-gdbserver cudaHashcat64.bin cuda-install-samples-9.0.sh cuda-memcheck
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ which cuda
cudafe cudafe++ cuda-gdb cuda-gdbserver cudaHashcat64.bin cuda-install-samples-9.0.sh cuda-memcheck
hp@hp-desktop:~/TensorRT-3.0.0/samples/sampleMNIST$ cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
#define CUDNN_MAJOR 7
#define CUDNN_MINOR 0
#define CUDNN_PATCHLEVEL 3
--
#define CUDNN_VERSION (CUDNN_MAJOR * 1000 + CUDNN_MINOR * 100 + CUDNN_PATCHLEVEL)
#include "driver_types.h"
cat .bashrc
export PATH=/usr/local/cuda-9.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64\${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
make CUDA_INSTALL_DIR=/usr/local/cuda-9/ CUDNN_INSTALL_DIR=/usr/local/cuda-9.0/
Compiling: sampleMNIST.cpp
Compiling: common.cpp
Linking: ../../bin/sample_mnist_debug
Compiling: sampleMNIST.cpp
Compiling: common.cpp
Linking: ../../bin/sample_mnist
ls -la /home/hp/TensorRT-3.0.0/bin/sample_mnist
-rwxrwxr-x 1 hp hp 643672 ott 19 13:39 /home/hp/TensorRT-3.0.0/bin/sample_mnist
sudo apt-get install nvinfer does not find anything on Ubuntu 16.04 64bit kernel 4.4.96. And I also cannot find nvdnn samples in the tar file, it seems only to be available with the the deb package. Should I reinstall tensorrt ?
Solved