./run.sh: line 78: th: command not found about alphapose

[Fang-Haoshu]

Hi, you need to install torch

[tedrickzhu]

I have installed torch ,I will check it again,thanks

[Fang-Haoshu]

maybe it's not installed correctly or not added to the bashrc file. If it's installed correctly, th command should be able to run it.

[zhijian-liu]

Or maybe you have just installed torch. In this case, you should do source ~/.bashrc to make th command available (or you can restart a Terminal session).

[tedrickzhu]

thanks ,it's ok ,but there is another problem ,init.lua:389: module 'hdf5' not found:No LuaRocks module found for hdf5, do I solve it with this? "luarocks install hdf5"

[Fang-Haoshu]

exactly

[tedrickzhu]

sorry,but I have another proplem,error as follow ,before all of these,I have this problem,"libcudnn.so.6: cannot open shared object file: No such file or directory",so I change my cudnn7.0 to 6.0,do I need to change it to cudnn5.0 now? I have installed cuda8.0,thanks

error infor:

home/zzy/torch/install/bin/luajit: /home/zzy/torch/install/share/lua/5.1/trepl/init.lua:389: /home/zzy/torch/install/share/lua/5.1/trepl/init.lua:389: /home/zzy/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"

[tedrickzhu]

I have added cudnn5.0 to cuda ,that's ok .but I want to know the device requirements ,because it's out of memery now, crying.....anyway ,thanks so much ,have a good day

[Fang-Haoshu]

how about cudnn7.0. We are using cudnn7.0 on our server

[Fang-Haoshu]

if it oom, try to add flag: --batch 1

[tedrickzhu]

using --batch 1 is ok ,thanks so much