![]() Python-libnvinfer : Depends: libnvinfer7 (= 7.0.0-1+cuda10.2) but 7.2.3-1+cuda11.1 is to be installed Libnvonnxparsers7 : Depends: libnvinfer7 (= 7.0.0-1+cuda10.2) but 7.2.3-1+cuda11.1 is to be installed ![]() The following packages have unmet dependencies: The following information may help to resolve the situation: Requested an impossible situation or if you are using the unstableĭistribution that some required packages have not yet been created I Tried **sudo apt install tensorrt libnvinfer7** bashrc export CUDA_HOME=/usr/local/cudaĮxport DYLD_LIBRARY_PATH=$CUDA_HOME/lib64:$DYLD_LIBRARY_PATHĮxport C_INCLUDE_PATH=$CUDA_HOME/include:$C_INCLUDE_PATHĮxport CPLUS_INCLUDE_PATH=$CUDA_HOME/include:$CPLUS_INCLUDE_PATHĮxport LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATHĮxport LD_RUN_PATH=$CUDA_HOME/lib64:$LD_RUN_PATH cat /usr/include/x86_64-linux-gnu/cudnn_v*.h | grep CUDNN_MAJOR -A 2ĭownload TensorRT here. This method of installation installs cuda in /usr/include and /usr/lib/cuda/lib64, hence the file you need to look at is in /usr/include/cudnn.h. Sudo mv cuda-ubuntu1804.pin /etc/apt/preferences.d/cuda-repository-pin-600ĭownload CuDNN here (BOTH the runtime and dev, deb) from: sudo dpkg -i libcudnn7_7.6.5.32-1+cuda10.2_b You can also install CUDA directly from the offline installer, but this is a little easier.
0 Comments
Leave a Reply. |