CuDNN v7 for Cuda 9.0 not working

[ProGamerGov]

I installed Cuda 9.0 and CuDNN v7, but I keep getting this error when trying to use Torch7's cudnn package.

/home/ubuntu/torch/install/bin/luajit: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/cudnn/ffi.lua:1603: 'libcudnn (R5) not found in library path.
Please install CuDNN from https://developer.nvidia.com/cuDNN
Then make sure files named as libcudnn.so.5 or libcudnn.5.dylib are placed in
your library load path (for example /usr/local/lib , or manually add a path to LD_LIBRARY_PATH)

Alternatively, set the path to libcudnn.so.5 or libcudnn.5.dylib
to the environment variable CUDNN_PATH and rerun torch.
For example: export CUDNN_PATH="/usr/local/cuda/lib64/libcudnn.so.5"

stack traceback:
        [C]: in function 'error'
        /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
        neural_style.lua:350: in function 'setup_gpu'
        neural_style.lua:53: in function 'main'
        neural_style.lua:601: in main chunk
        [C]: in function 'dofile'
        ...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50
ubuntu@ip-Address:~/neural-style$

[ProGamerGov]

Using this: export CUDNN_PATH="/usr/local/cuda-9.0/lib64/libcudnn.so.7"

Results in:

Found Environment variable CUDNN_PATH = /usr/local/cuda-9.0/lib64/libcudnn.so.7/home/ubuntu/torch/install/bin/luajit: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/cudnn/ffi.lua:1618: These bindings are for CUDNN 5.x (5005 <= cudnn.version > 6000) , while the loaded CuDNN is version: 7003
Are you using an older or newer version of CuDNN?
stack traceback:
        [C]: in function 'error'
        /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
        neural_style.lua:350: in function 'setup_gpu'
        neural_style.lua:53: in function 'main'
        neural_style.lua:601: in main chunk
        [C]: in function 'dofile'
        ...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50

Edit:

Using this worked:

git clone https://github.com/soumith/cudnn.torch.git -b R7 && cd cudnn.torch && luarocks make cudnn-scm-1.rockspec

[soulslicer]

got to do so much bullshit to make torch/lua work with 10 series nvidia GPU. even after 1 year +.

[MozhdehDokhani]

In which path did you run the below code?!
git clone https://github.com/soumith/cudnn.torch.git -b R7 && cd cudnn.torch && luarocks make cudnn-scm-1.rockspec

[MozhdehDokhani]

I have the same problem and I couldnt solve it yet!
I successfully installed my nvidia driver and cuda 9.1. I also did everything that is needed to install cudnn 7.1. but when I run my code, I face the following error:

/////////////////////////////////////////////////////////////////
Found Environment variable CUDNN_PATH = /usr/local/cuda-9.1/lib64/libcudnn.so.7/home/mozhdeh/torch/install/bin/luajit: /home/mozhdeh/torch/install/share/lua/5.1/trepl/init.lua:389: /home/mozhdeh/torch/install/share/lua/5.1/trepl/init.lua:389: /home/mozhdeh/torch/install/share/lua/5.1/cudnn/ffi.lua:1743: /usr/local/cuda-9.1/lib64/libcudnn.so.7: cannot open shared object file: No such file or directory
/////////////////////////////////////////////////////////////////

My .bashrc file containes:

/////////////////////////////////////////////////////////////////
export PATH=/usr/local/cuda-9.1/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-9.1/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
export CUDA_HOME=/usr/local/cuda-9.1
export CUDNN_PATH=/usr/local/cuda-9.1/lib64/libcudnn.so.7
/////////////////////////////////////////////////////////////////

and I'm sure that the libcudnn.so.7 file exists in /usr/local/cuda-9.1/lib64
Any solution ?! :(

System Information:
Os: Ubuntu 18.04
Nvidia driver version: 390.48

[Montroigenc]

You can change the called cudnn version executing the following command in the prompt:

export CUDNN_PATH="/usr/local/cuda/lib64/libcudnn.so.X"

where the X is your current cudnn version, in my case version 7.

[ShuaiHuang]

Using this: export CUDNN_PATH="/usr/local/cuda-9.0/lib64/libcudnn.so.7"

Results in:

Found Environment variable CUDNN_PATH = /usr/local/cuda-9.0/lib64/libcudnn.so.7/home/ubuntu/torch/install/bin/luajit: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/cudnn/ffi.lua:1618: These bindings are for CUDNN 5.x (5005 <= cudnn.version > 6000) , while the loaded CuDNN is version: 7003
Are you using an older or newer version of CuDNN?
stack traceback:
        [C]: in function 'error'
        /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
        neural_style.lua:350: in function 'setup_gpu'
        neural_style.lua:53: in function 'main'
        neural_style.lua:601: in main chunk
        [C]: in function 'dofile'
        ...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50

Edit:

Using this worked:

git clone https://github.com/soumith/cudnn.torch.git -b R7 && cd cudnn.torch && luarocks make cudnn-scm-1.rockspec

THX, it works for me!

[mememimis]

Using this: export CUDNN_PATH="/usr/local/cuda-9.0/lib64/libcudnn.so.7"
Results in:

Found Environment variable CUDNN_PATH = /usr/local/cuda-9.0/lib64/libcudnn.so.7/home/ubuntu/torch/install/bin/luajit: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/cudnn/ffi.lua:1618: These bindings are for CUDNN 5.x (5005 <= cudnn.version > 6000) , while the loaded CuDNN is version: 7003
Are you using an older or newer version of CuDNN?
stack traceback:
        [C]: in function 'error'
        /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
        neural_style.lua:350: in function 'setup_gpu'
        neural_style.lua:53: in function 'main'
        neural_style.lua:601: in main chunk
        [C]: in function 'dofile'
        ...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50

Edit:
Using this worked:
git clone https://github.com/soumith/cudnn.torch.git -b R7 && cd cudnn.torch && luarocks make cudnn-scm-1.rockspec

THX, it works for me!

This does not work for me. Im on ubuntu 18.02 cuda 9.2 and cudnn 7

Any ideas?

[THUHoloLab]

Using this: export CUDNN_PATH="/usr/local/cuda-9.0/lib64/libcudnn.so.7"
Results in:

Found Environment variable CUDNN_PATH = /usr/local/cuda-9.0/lib64/libcudnn.so.7/home/ubuntu/torch/install/bin/luajit: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: /home/ubuntu/torch/install/share/lua/5.1/cudnn/ffi.lua:1618: These bindings are for CUDNN 5.x (5005 <= cudnn.version > 6000) , while the loaded CuDNN is version: 7003
Are you using an older or newer version of CuDNN?
stack traceback:
        [C]: in function 'error'
        /home/ubuntu/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
        neural_style.lua:350: in function 'setup_gpu'
        neural_style.lua:53: in function 'main'
        neural_style.lua:601: in main chunk
        [C]: in function 'dofile'
        ...untu/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
        [C]: at 0x00405d50

Edit:
Using this worked:
git clone https://github.com/soumith/cudnn.torch.git -b R7 && cd cudnn.torch && luarocks make cudnn-scm-1.rockspec

THX, it works for me!

This does not work for me. Im on ubuntu 18.02 cuda 9.2 and cudnn 7

Any ideas?

It also does not work for me. Im on ubuntu 16.04 cuda 9.0 and cudnn 7.4.2

Are you using an older or newer version of CuDNN?
stack traceback:
	[C]: in function 'error'
	/home/user/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
	TestRDN.lua:3: in main chunk
	[C]: in function 'dofile'
	...user/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
	[C]: at 0x00405d50

[djiajunustc]

@THUHoloLab
Hi,
I have met the same problem as yours.
Have you solved it?

Thanks