Closed liuquan0722 closed 8 years ago
hughperkins
commented
8 years ago Hi Jimmy,
Hmmm, I could reproduce this, and then I did luarocks install loadcaffe, and the problem disappeared. Can you try doing luarocks install loadcaffe, and see if that changes something please?
liuquan0722
commented
8 years ago I reinstalled loadcoffe. Kindly view below log. But the issue appears. :-( Bad luck to me
quans-Mac-mini:neural-style liuquan0722$ luarocks install loadcaffe
Installing https://raw.githubusercontent.com/torch/rocks/master/loadcaffe-1.0-0.rockspec...
Using https://raw.githubusercontent.com/torch/rocks/master/loadcaffe-1.0-0.rockspec... switching to 'build' mode
Cloning into 'loadcaffe'...
remote: Counting objects: 18, done.
remote: Compressing objects: 100% (17/17), done.
remote: Total 18 (delta 0), reused 6 (delta 0), pack-reused 0
Receiving objects: 100% (18/18), 18.81 KiB | 0 bytes/s, done.
Checking connectivity... done.
cmake -E make_directory build;
cd build;
cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/Users/liuquan0722/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0";
make
-- The C compiler identification is AppleClang 7.0.0.7000176
-- The CXX compiler identification is AppleClang 7.0.0.7000176
-- Check for working C compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc
-- Check for working C compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++
-- Check for working CXX compiler: /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found Torch7 in /Users/liuquan0722/torch/install
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - found
-- Found Threads: TRUE
-- Found PROTOBUF: /usr/local/lib/libprotobuf.dylib
-- Configuring done
-- Generating done
-- Build files have been written to: /tmp/luarocks_loadcaffe-1.0-0-48/loadcaffe/build
[ 25%] Running C++ protocol buffer compiler on caffe.proto
Scanning dependencies of target loadcaffe
[ 50%] Building CXX object CMakeFiles/loadcaffe.dir/loadcaffe.cpp.o
[ 75%] Building CXX object CMakeFiles/loadcaffe.dir/caffe.pb.cc.o
[100%] Linking CXX shared module libloadcaffe.so
[100%] Built target loadcaffe
cd build && make install
[100%] Built target loadcaffe
Install the project...
-- Install configuration: "Release"
-- Installing: /Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0/lib/libloadcaffe.so
-- Installing: /Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0/lua/loadcaffe/ffi.lua
-- Installing: /Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0/lua/loadcaffe/init.lua
-- Installing: /Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0/lua/loadcaffe/loadcaffe.lua
-- Installing: /Users/liuquan0722/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0/lua/loadcaffe/test.lua
Updating manifest for /Users/liuquan0722/torch/install/lib/luarocks/rocks
loadcaffe 1.0-0 is now built and installed in /Users/liuquan0722/torch/install/ (license: BSD)run neural_style.lua again
/Users/liuquan0722/torch/install/bin/luajit: ...22/torch/install/share/lua/5.1/nn/SpatialConvolution.lua:100: bad argument #2 to 'SpatialConvolutionMM_updateOutput' (input channels and nInputPlane dont match)
stack traceback:
[C]: in function 'SpatialConvolutionMM_updateOutput'
...22/torch/install/share/lua/5.1/nn/SpatialConvolution.lua:100: in function 'updateOutput'
...iuquan0722/torch/install/share/lua/5.1/nn/Sequential.lua:44: in function 'forward'
neural_style.lua:176: in function 'main'
neural_style.lua:497: in main chunk
[C]: in function 'dofile'
...0722/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x010749cbc0Hmmm, that's odd. I think it'd be good to make sure everything is a recent version. Can you do the following please:
luarocks install torch7
luarocks install nn
luarocks install cltorch
luarocks install clnn
luajit -l cltorch -e 'cltorch.about()'
luajit -l clnn -e 'clnn.about()'... and provide the output from the last couple of commands, and also rerun the neural_style, and check what happens. And also provide the output from git log -n 5 --oneline, from within the neural-style directory.
liuquan0722
commented
8 years ago Hi Hugh, Sorry for late response.
As per your suggestion, I installed them as below: 1, install itorch as per the instruction http://torch.ch/docs/getting-started.html#_ 2, resolved ipython not found issue. https://github.com/facebook/iTorch/issues/5 3, install nn 4, install cltorch 5, install clnn 6 luajit -l cltorch -e 'cltorch.about()' 7 luajit -l clnn -e 'clnn.about()'
Output:
(ip)quans-Mac-mini:neural-style liuquan0722$ luajit -l cltorch -e 'cltorch.about()'
cltorch. OpenCL backend for Torch
Built from commit ddb1964
More info, doc: https://github.com/hughperkins/cltorch
Issues: https://github.com/hughperkins/cltorch/issues
(ip)quans-Mac-mini:neural-style liuquan0722$ luajit -l clnn -e 'clnn.about()'
libthclnn_searchpath /Users/liuquan0722/torch/install/lib/lua/5.1/libTHCLNN.so
clnn. OpenCL backend for Torch nn
Built from commit
More info, doc: https://github.com/hughperkins/clnn
Issues: https://github.com/hughperkins/clnn/issues
(ip)quans-Mac-mini:neural-style liuquan0722$ git log -n 5 --oneline
8b7102f Merged from jcjohnson/neural-style master branch
6749437 OpenCL Support complete.
9e0b57d Merge branch 'master' of https://github.com/jcjohnson/neural-style
8795bbf fix loadcaffe_wrapper to work again in CPU mode
42f83bc Merge pull request #54 from bcipolli/resume_downloadsMany thanks for your patience on the issue. I still get the same error.
(ip)quans-Mac-mini:neural-style liuquan0722$ 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
libthclnn_searchpath /Users/liuquan0722/torch/install/lib/lua/5.1/libTHCLNN.so
Successfully loaded models/nin_imagenet_conv.caffemodel
MODULE data UNDEFINED
warning: module 'data [type 5]' not found
Changing line: table.insert(model, {'pool4', nn.SpatialAveragePooling(6, 6, 1, 1):ceil()})
To line: table.insert(model, {'pool4', nn.SpatialAveragePooling(6, 6, 1, 1):ceil():ceil()})
cccp1: 96 96 1 1
cccp2: 96 96 1 1
conv2: 256 96 5 5
cccp3: 256 256 1 1
cccp4: 256 256 1 1
conv3: 384 256 3 3
cccp5: 384 384 1 1
cccp6: 384 384 1 1
conv4-1024: 1024 384 3 3
cccp7-1024: 1024 1024 1 1
cccp8-1024: 1000 1024 1 1
Using Apple , OpenCL platform: Apple
Using OpenCL device: ATI Radeon HD 6630M
Setting up content layer 1 : relu0
Setting up style layer 1 : relu0
WARNING: Skipping content loss
Setting up content layer 8 : relu3
THClReduceAll.cl build log:
<program source>:11:10: warning: unused variable 'in1'
float *in1 = &_in1;
^
<program source>:12:10: warning: unused variable 'out'
float *out = &_out;
^
/Users/liuquan0722/torch/install/bin/luajit: ...22/torch/install/share/lua/5.1/nn/SpatialConvolution.lua:100: bad argument #2 to 'SpatialConvolutionMM_updateOutput' (input channels and nInputPlane dont match)
stack traceback:
[C]: in function 'SpatialConvolutionMM_updateOutput'
...22/torch/install/share/lua/5.1/nn/SpatialConvolution.lua:100: in function 'updateOutput'
...iuquan0722/torch/install/share/lua/5.1/nn/Sequential.lua:44: in function 'forward'
neural_style.lua:176: in function 'main'
neural_style.lua:497: in main chunk
[C]: in function 'dofile'
...0722/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x0106639bc0Torch7 should be installed as well
(ip)quans-Mac-mini:neural-style liuquan0722$ th
______ __ | Torch7
/_ __/__ ________/ / | Scientific computing for Lua.
/ / / _ \/ __/ __/ _ \ | Type ? for help
/_/ \___/_/ \__/_//_/ | https://github.com/torch
| http://torch.ch
th> torch.Tensor{1,2,3}
1
2
3
[torch.DoubleTensor of size 3]
[0.0315s]
th> Hmmm, I probably need to provide some kind of debug version, so I can see what is going on. Will try to find a moment to do this.
hughperkins
commented
8 years ago Interestingly, my output is quite different:
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
Successfully loaded models/nin_imagenet_conv.caffemodel
conv1: 96 3 11 11
cccp1: 96 96 1 1
cccp2: 96 96 1 1
conv2: 256 96 5 5
cccp3: 256 256 1 1
cccp4: 256 256 1 1
conv3: 384 256 3 3
cccp5: 384 384 1 1
cccp6: 384 384 1 1
conv4-1024: 1024 384 3 3
cccp7-1024: 1024 1024 1 1
cccp8-1024: 1000 1024 1 1
Using NVIDIA Corporation , OpenCL platform: NVIDIA CUDA
Using OpenCL device: GeForce 940M
Setting up content layer 2 : relu0
Setting up style layer 2 : relu0
WARNING: Skipping content loss
Setting up content layer 9 : relu3
Setting up style layer 9 : relu3
WARNING: Skipping content loss
Setting up content layer 16 : relu7
Setting up style layer 16 : relu7
WARNING: Skipping content loss
Setting up content layer 28 : relu12
Setting up style layer 28 : relu12 Look at the numbers eg of the first line. You get, well, you're missing the conv1 layer for some reason...
hughperkins
commented
8 years ago Can you provide the output of the following please:
luajit -l loadcaffe -e "loadcaffe.load('models/train_val.prototxt', 'models/nin_imagenet_conv.caffemodel', 'nn')"Also, can you do luarocks install loadcaffe, and retry, and confirm output after reinstalling loadcaffe? (Also, can you paste the output from running luarocks install loadcaffe please?)
hughperkins
commented
8 years ago Since this is kind of old, closing it. Feel free to reopen it if you still need it.
Hi Hugh, When I run the neural-style.lua with openCL, I got below errors. it seems some library is missing, right? Appreciate it if the issue could be looked at.