Can't load libtorch.so

[bhive01]

Saw this and loved the idea. Downloaded and installed all of the required software (took a while!), but torch seems to be working fine. Installed the OpenCL part and downloaded the NiN model to run your demo, but I get an error.

Brandons-iMac:neural-style brandonhurr$ th neural_style.lua -style_image examples/inputs/picasso_selfport1907.jpg -content_image examples/inputs/brad_pitt.jpg -output_image profile.png -model_file models/nin_imagenet_conv.caffemodel -proto_file models/train_val.prototxt -gpu 0 -backend clnn -num_iterations 1000 -seed 123 -content_layers relu0,relu3,relu7,relu12 -style_layers relu0,relu3,relu7,relu12 -content_weight 10 -style_weight 1000 -image_size 512 -optimizer adam
/Users/brandonhurr/torch/install/bin/luajit: ...s/brandonhurr/torch/install/share/lua/5.1/trepl/init.lua:383: ...s/brandonhurr/torch/install/share/lua/5.1/trepl/init.lua:383: ...brandonhurr/torch/install/share/lua/5.1/cltorch/init.lua:19: cannot load '/Users/brandonhurr/torch/install/lib/lua/5.1/libcltorch.so'
stack traceback:
    [C]: in function 'error'
    ...s/brandonhurr/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
    neural_style.lua:55: in function 'main'
    neural_style.lua:499: in main chunk
    [C]: in function 'dofile'
    ...hurr/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
    [C]: at 0x01010fbbc0

I verified that libtorch.so exists so I'm unsure why I can't get past this point.

Brandons-iMac:neural-style brandonhurr$ luajit -l clnn -e 'clnn.test()'
libthclnn_searchpath    /Users/brandonhurr/torch/install/lib/lua/5.1/libTHCLNN.so
Running 64 tests
|_______________________________________________________________  ==> Abs_backwardUsing Apple , OpenCL platform: Apple
Using OpenCL device: Iris Pro
____|___________________________________________________________  ==> ClassNLLCriterionMultipleTargetTHClReduceAll.cl build log: 
<program source>:9:10: warning: unused variable 'in1'
  float *in1 = &_in1;
         ^
<program source>:10:10: warning: unused variable 'out'
  float *out = &_out;
         ^

_____________________________________________________________|__  ==> mseApply_3t_0s_0pt_-2_-2_-2_*out = 0.00043487714720591 * (*in1 - *in2) build log: 
<program source>:37:12: warning: double precision constant requires cl_khr_fp64, casting to single precision
    *out = 0.00043487714720591 * (*in1 - *in2);
           ^

______________________________________________________________|_  ==> mse_nosizeaverageApply_3t_0s_0pt_-2_-2_-2_*out = 0.00040675208460443 * (*in1 - *in2) build log: 
<program source>:37:12: warning: double precision constant requires cl_khr_fp64, casting to single precision
    *out = 0.00040675208460443 * (*in1 - *in2);
           ^

________________________________________________________________  ==> Done                 

Completed 98 asserts in 64 tests with 0 errors

--------------------------------------------------------------------------------

[tigerneil]

@bhive01 I still cannot get this work in th.(in LuaJIT no problem)

LuaJIT 2.1.0-beta1 -- Copyright (C) 2005-2015 Mike Pall. http://luajit.org/

 _____              _
|_   _|            | |
  | | ___  _ __ ___| |__
  | |/ _ \| '__/ __| '_ \
  | | (_) | | | (__| | | |
  \_/\___/|_|  \___|_| |_|

JIT: ON SSE2 SSE3 SSE4.1 fold cse dce fwd dse narrow loop abc sink fuse
th> require 'nn'
th> require 'clnn'
libthclnn_searchpath    /Users/zhuxiaohu/torch/install/lib/lua/5.1/libTHCLNN.so
th> require 'cltorch'
th> exit
th>> ^C

have u finally solved this problem?

[bhive01]

Yes, Apologies for not linking to the solution.

I got with someone on libtorch.
hughperkins/cltorch#24 (comment)

Found another message on the same library that suggested deleting #bin/bash from a single script file. Then it worked.
hughperkins/cltorch#21

See here on what to do. It works pretty well and is a lot of fun to play with. It looks like they might have pinned it down to another library TREPL.
torch/trepl#32

[tigerneil]

thanks @bhive01 I found the problem is the usage of zsh shell. After I change to bash shell, everything goes OK. Hope all zsh guys knew this situation.

[markostam]

deleting the #!bin/bash worked for me as well on el capitan thank you.