./run.sh failed!

[xtttttttttx]

I've installed cuda9.0 and cudnn v7. What is *.raw?

▶ ./run.sh                                                 
/home/artprog/usr/local/torch/install/bin/luajit: cannot open main.lua: No such file or directory
stack traceback:
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
ls: cannot access '*.txt': No such file or directory
Found Environment variable CUDNN_PATH = /usr/lib/x86_64-linux-gnu/libcudnn.so7/home/artprog/usr/local/torch/install/bin/luajit: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...prog/usr/local/torch/install/share/lua/5.1/cudnn/ffi.lua:1592: /usr/lib/x86_64-linux-gnu/libcudnn.so7: cannot open shared object file: No such file or directory
stack traceback:
    [C]: in function 'error'
    ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
    process.lua:17: in main chunk
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
ls: cannot access '*.raw': No such file or directory

I've tried to run from matlab. But it complained as follows:

device =

gpu

zsh:1: command not found: th
Error using run (line 38)
Failed to run Torch7 script.

But torch works in terminal.

[AaronJackson]

Hi, did you clone the git repo recursively? it looks like I forgot to add the --recursive flag to the vrn git clone. I will correct this immediately.

[xtttttttttx]

still the same error.

[AaronJackson]

Can you please be more specific? In your first post, the first error was that it was unable to find main.lua. if you cloned recursively, this file will exist under face-alignment/main.lua.

[xtttttttttx]

@AaronJackson The folder face-alignment does contain the main.lua. Now I run it in matlab, it complained like this:

device =

gpu

/home/artprog/usr/local/torch/install/bin/luajit: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...sr/local/torch/install/share/lua/5.1/luarocks/loader.lua:117: error loading module 'fb.python.lib' from file '/home/artprog/usr/local/torch/install/lib/lua/5.1/fb/python/lib.so':
    /usr/local/MATLAB/R2015b/sys/os/glnxa64/libstdc++.so.6: version `CXXABI_1.3.8' not found (required by /home/artprog/usr/local/torch/install/lib/lua/5.1/fb/python/lib.so)
stack traceback:
    [C]: in function 'error'
    ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
    main.lua:8: in main chunk
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
Error using run (line 38)
Failed to run Torch7 script.

[AaronJackson]

I believe the matlab error might be due to an old version of matlab. One quick fix for this might be to start matlab like so:

LD_PRELOAD_PATH=/usr/lib64/libstdc++.so.6 matlab

Can you also test the run.sh script.

[xtttttttttx]

Matlab still says:

device =

gpu

/home/artprog/usr/local/torch/install/bin/luajit: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...sr/local/torch/install/share/lua/5.1/luarocks/loader.lua:117: error loading module 'fb.python.lib' from file '/home/artprog/usr/local/torch/install/lib/lua/5.1/fb/python/lib.so':
    /usr/local/MATLAB/R2015b/sys/os/glnxa64/libstdc++.so.6: version `CXXABI_1.3.8' not found (required by /home/artprog/usr/local/torch/install/lib/lua/5.1/fb/python/lib.so)
stack traceback:
    [C]: in function 'error'
    ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
    main.lua:8: in main chunk
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
Error using run (line 38)
Failed to run Torch7 script.

run.sh complained like this:

▶ ./run.sh 
Found Environment variable CUDNN_PATH = /usr/lib/x86_64-linux-gnu/libcudnn.so.7/home/artprog/usr/local/torch/install/bin/luajit: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...prog/usr/local/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: 7002  
Are you using an older or newer version of CuDNN?
stack traceback:
    [C]: in function 'error'
    ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
    main.lua:13: in main chunk
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
ls: cannot access '*.txt': No such file or directory
Found Environment variable CUDNN_PATH = /usr/lib/x86_64-linux-gnu/libcudnn.so.7/home/artprog/usr/local/torch/install/bin/luajit: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: ...prog/usr/local/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: 7002  
Are you using an older or newer version of CuDNN?
stack traceback:
    [C]: in function 'error'
    ...rog/usr/local/torch/install/share/lua/5.1/trepl/init.lua:389: in function 'require'
    process.lua:17: in main chunk
    [C]: in function 'dofile'
    ...ocal/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:150: in main chunk
    [C]: at 0x00405d50
ls: cannot access '*.raw': No such file or directory

Do I need reinstall cudnn of version 5?

[AaronJackson]

Which version of Linux are you using? Can you run find /usr -name "libstdc++.so.6" to find the correct path to find your libstdc, and then use that path with the LD_PRELOAD_PATH environment variable?

The version of CuDNN might be an issue.

[AaronJackson]

Ah, it should be LD_PRELOAD not LD_PRELOAD_PATH

[1adrianb]

@TianxinTse Cudnn 7.0 support is experimental as far as I know. From what I can see you are using the wrong cudnn package, please make sure you are using the R7 branch https://github.com/soumith/cudnn.torch/tree/R7 or better, use cudnn5 :)

[xtttttttttx]

@AaronJackson @1adrianb Thank you for you guys' help and advice. I will try cudnn5.