TensorFlow 模型服务器 GPU 构建 - cudnn 路径
TensorFlow Model Server GPU Build - cudnn path
我正在尝试在 RedHat 计算机上从源构建 TensorFlow 模型服务器。我没有 sudo 权限,cudnn 不在默认目录中。有没有办法在构建时指定cudnn路径?
Cuda Configuration Error: Failed to run find_cuda_config.py: Could not find any cudnn.h matching version '10.0' in any subdirectory:
''
'include'
'include/cuda'
'include/*-linux-gnu'
'extras/CUPTI/include'
'include/cuda/CUPTI'
of:
'/lib'
'/lib64'
'/opt/beegfs/lib'
'/usr'
'/usr/lib64/atlas'
'/usr/lib64/dyninst'
'/usr/lib64/mysql'
'/usr/lib64/tcl8.5'
'/usr/local/cuda'
'/usr/local/cuda-10.0/targets/x86_64-linux/lib'
'/usr/local/cuda-8.0/targets/x86_64-linux/lib'
'/usr/local/cuda-9.0/targets/x86_64-linux/lib'
'/usr/local/cuda-9.1/targets/x86_64-linux/lib'
为了社区的利益,即使它出现在评论部分中,也在此(答案)部分中提供解决方案。
我们需要在运行 Bazel
,
之前使用以下命令指定CUDNN路径
export CUDNN_INSTALL_PATH=<...>
我正在尝试在 RedHat 计算机上从源构建 TensorFlow 模型服务器。我没有 sudo 权限,cudnn 不在默认目录中。有没有办法在构建时指定cudnn路径?
Cuda Configuration Error: Failed to run find_cuda_config.py: Could not find any cudnn.h matching version '10.0' in any subdirectory:
''
'include'
'include/cuda'
'include/*-linux-gnu'
'extras/CUPTI/include'
'include/cuda/CUPTI'
of:
'/lib'
'/lib64'
'/opt/beegfs/lib'
'/usr'
'/usr/lib64/atlas'
'/usr/lib64/dyninst'
'/usr/lib64/mysql'
'/usr/lib64/tcl8.5'
'/usr/local/cuda'
'/usr/local/cuda-10.0/targets/x86_64-linux/lib'
'/usr/local/cuda-8.0/targets/x86_64-linux/lib'
'/usr/local/cuda-9.0/targets/x86_64-linux/lib'
'/usr/local/cuda-9.1/targets/x86_64-linux/lib'
为了社区的利益,即使它出现在评论部分中,也在此(答案)部分中提供解决方案。
我们需要在运行 Bazel
,
export CUDNN_INSTALL_PATH=<...>