Comments (40)
You should be able to just luarocks install hdf5
from neuraltalk2.
Also, it's unfortunate that so many dependencies are needed. I wonder if it makes sense to create a thin container (e.g. in docker) that has everything installed and can run a CPU forward? Something I might look into over the weekend.
from neuraltalk2.
yay! Thank you @PaulSchnau and @soumith for pitching in - I'm sure this should be helpful to many others who might follow in your path. Closing the issue.
from neuraltalk2.
@jackyliang Try brew install hdf5
from neuraltalk2.
@jackyliang now run luarocks install hdf5
from neuraltalk2.
Okay, wait a second.
I guess, I forgot to download the actual NeuralTalk repository?
So, after I do this, will I just have to put this entire folder somewhere (anywhere?) and when putting $ th eval.lua -model /path/to/model -image_folder /path/to/image/directory -num_images 10
in the Terminal I simply have to point it to the actual path of the eval.lua file? Or does it work differently?
Because when I now try to do that, then it complains that I don't have loadcaffe installed.
/Users/username/torch/install/bin/luajit: /Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: module 'loadcaffe' not found:No LuaRocks module found for loadcaffe
However, in the readme it states that you only need loadcaffe if you want to train your models (which I don't want). If I try to install loadcaffe it gives me this error after a while:
-- Could NOT find PROTOBUF (missing: PROTOBUF_LIBRARY PROTOBUF_INCLUDE_DIR)
CMake Error: The following variables are used in this project, but they are set to NOTFOUND.
Please set them or make sure they are set and tested correctly in the CMake files:
PROTOBUF_LIBRARY (ADVANCED)
linked by target "loadcaffe" in directory /tmp/luarocks_loadcaffe-1.0-0-9224/loadcaffe
-- Configuring incomplete, errors occurred!
See also "/tmp/luarocks_loadcaffe-1.0-0-9224/loadcaffe/build/CMakeFiles/CMakeOutput.log".
make: *** No targets specified and no makefile found. Stop.
Error: Build error: Failed building.
Anybody knows what I am doing wrong here? Please explain in noobish terms. ;)
from neuraltalk2.
brew install protobuf
from neuraltalk2.
Ok, now I am very confused. I did the brew install protobuf
that you suggested, and now when I want to install loadcaffe again it tells me -bash: luarocks: command not found
as well as -bash: th: command not found
(I had restarted the Terminal in the meantime). Do I need to "activate" torch somehow? Because it did work earlier...
from neuraltalk2.
Follow the instructions for installing Torch from http://torch.ch/docs/getting-started.html. You should have lua, luarocks, and th when you are done.
from neuraltalk2.
PaulSchnau, thanks for helping me.
That's the instructions I followed.
And luarocks and th did work earlier. Now after restarting they do not work anymore.
I now redid the last step of the installation ( cd ~/torch; ./install.sh
) and in the end it tells me
Not updating your shell profile.
You might want to
add the following lines to your shell profile:
. /Users/username/torch/install/bin/torch-activate
I guess that's what I was missing the first time around. But how do I add this line to "my shell profile", what is my shell profile, where do I find it?
from neuraltalk2.
Your shell (bash) profile is a set of commands run when you start up bash. Adding that line to your profile will activate torch when you run a th command.
You can manually edit your bash profile using a text editor like vim or nano:
sudo vim ~/.profile
sudo nano ~/.profile
And paste that line in (replacing username with your real username of course). Then you can reload your profile using the command from the Torch install instructions.
source ~/.profile
th commands should then work.
from neuraltalk2.
This works. Although I seem to need to run source ~/.profile
every time after I restart the Terminal.
I also successfully installed protobuf, however installing loadcaffe still does not work. It reaches 50% and then stops with this:
[ 50%] Building CXX object CMakeFiles/loadcaffe.dir/loadcaffe.cpp.o
/tmp/luarocks_loadcaffe-1.0-0-8793/loadcaffe/loadcaffe.cpp:152:17: error: no member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char> > > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-8793/loadcaffe/loadcaffe.cpp:164:17: error: no member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char> > > >'
lines.emplace_back(layer.name(), buf);
[then 18 more of those...]
fatal error: too many errors emitted, stopping now [-ferror-limit=]
20 errors generated.
make[2]: *** [CMakeFiles/loadcaffe.dir/loadcaffe.cpp.o] Error 1
make[1]: *** [CMakeFiles/loadcaffe.dir/all] Error 2
make: *** [all] Error 2
What is happening here?
from neuraltalk2.
Looks like the same issue as szagoruyko/loadcaffe#33
Update your gcc:
brew install gcc
and then try to install loadcaffe again.
from neuraltalk2.
Paul, thanks for bearing with me.
I did that, but still get the same error. Here is the complete "log", in case that helps somehow.
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 7 (delta 0), pack-reused 0
Receiving objects: 100% (18/18), 18.65 KiB | 0 bytes/s, done.
Checking connectivity... done.
cmake -E make_directory build;
cd build;
cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/Users/username/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/Users/username/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0";
make
-- The C compiler identification is AppleClang 5.1.0.5030040
-- The CXX compiler identification is AppleClang 5.1.0.5030040
-- 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/username/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-1.0-0-2008/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
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:152:17: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:164:17: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:171:17: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:228:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:236:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "cudnn.ReLU(true)");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:239:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "nn.ReLU(true)");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:249:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "cudnn.Tanh(true)");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:252:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "nn.Tanh()");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:262:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "cudnn.Sigmoid(true)");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:265:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "nn.Sigmoid()");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:282:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:295:19: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back("torch_transpose_bdwh", "nn.Transpose({4,...
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:296:17: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back("torch_view", "nn.View(-1):setNumInputDims(3)");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:298:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:306:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:311:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "nn.SoftMax()");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:316:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), "nn.SoftMax()");
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:400:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
/tmp/luarocks_loadcaffe-1.0-0-2008/loadcaffe/loadcaffe.cpp:412:15: error: no
member named 'emplace_back' in
'std::vector<std::pair<std::basic_string<char>, std::basic_string<char> >,
std::allocator<std::pair<std::basic_string<char>, std::basic_string<char>
> > >'
lines.emplace_back(layer.name(), buf);
~~~~~ ^
fatal error: too many errors emitted, stopping now [-ferror-limit=]
20 errors generated.
make[2]: *** [CMakeFiles/loadcaffe.dir/loadcaffe.cpp.o] Error 1
make[1]: *** [CMakeFiles/loadcaffe.dir/all] Error 2
make: *** [all] Error 2
Are there any other information I could provide to find out what is going wrong?
Also, am I correct in assuming that contrary to the NeuralTalk readme I actually really need loadcaffe to be installed?
from neuraltalk2.
I don't know if it will work, but I guess try what they did:
export CC=gcc-4.8
export CXX=g++-4.8
And then install loadcaffe again.
If that doesn't work, you can try getting the CUDA toolkit, even though you won't be able to use it.
https://developer.nvidia.com/cuda-toolkit
from neuraltalk2.
You can also try removing the require 'loadcaffe' from eval.lua, but I think some of the Tensor models require it.
from neuraltalk2.
Ok, after the export commands, I'm getting this error:
username$ 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 7 (delta 0), pack-reused 0
Receiving objects: 100% (18/18), 18.65 KiB | 0 bytes/s, done.
Checking connectivity... done.
cmake -E make_directory build;
cd build;
cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/Users/username/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/Users/username/torch/install/lib/luarocks/rocks/loadcaffe/1.0-0";
make
CMake Error at /usr/local/Cellar/cmake/3.4.0/share/cmake/Modules/CMakeDetermineCCompiler.cmake:57 (message):
Could not find compiler set in environment variable CC:
gcc-4.8.
Call Stack (most recent call first):
CMakeLists.txt
CMake Error: CMAKE_C_COMPILER not set, after EnableLanguage
CMake Error: CMAKE_CXX_COMPILER not set, after EnableLanguage
-- Configuring incomplete, errors occurred!
See also "/tmp/luarocks_loadcaffe-1.0-0-8118/loadcaffe/build/CMakeFiles/CMakeOutput.log".
make: *** No targets specified and no makefile found. Stop.
Error: Build error: Failed building.
from neuraltalk2.
We need to install that specific version of gcc:
brew install gcc48
from neuraltalk2.
This worked! So, now I have successfully installed loadcaffe.
However, when I run the NeuralTalk command, I still get an error:
/Users/username/torch/install/bin/luajit: /Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: module 'misc.utils' not found:No LuaRocks module found for misc.utils
no field package.preload['misc.utils']
no file '/Users/username/.luarocks/share/lua/5.1/misc/utils.lua'
no file '/Users/username/.luarocks/share/lua/5.1/misc/utils/init.lua'
no file '/Users/username/torch/install/share/lua/5.1/misc/utils.lua'
no file '/Users/username/torch/install/share/lua/5.1/misc/utils/init.lua'
no file './misc/utils.lua'
no file '/Users/username/torch/install/share/luajit-2.1.0-alpha/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils/init.lua'
no file '/Users/username/.luarocks/lib/lua/5.1/misc/utils.so'
no file '/Users/username/torch/install/lib/lua/5.1/misc/utils.so'
no file './misc/utils.so'
no file '/usr/local/lib/lua/5.1/misc/utils.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
no file '/Users/username/.luarocks/lib/lua/5.1/misc.so'
no file '/Users/username/torch/install/lib/lua/5.1/misc.so'
no file './misc.so'
no file '/usr/local/lib/lua/5.1/misc.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
stack traceback:
[C]: in function 'error'
/Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
/Users/username/Downloads/neuraltalk2-master/eval.lua:7: in main chunk
[C]: in function 'dofile'
...username/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x010de83f10
Do I do this correctly actually? I download the entire neuraltalk2 repository, place it whereever (?) and then in the command I provide the entire path to eval.lua? Correct?
from neuraltalk2.
Cool. You just need to add -gpu -1
to your command because you're not using a GPU.
from neuraltalk2.
(I figure you mean -gpuid -1)
Ok, I feel like we are getting really close, but it still does not quite work.
I downloaded the CPU model, but I still get the same error.
Literally my command is
$ th /Users/username/Desktop/neuraltalk2-master/eval.lua -model /Users/username/Desktop/checkpoint_v1.t7_cpu.t7 -image_folder /Users/username/Desktop/images -num_images 10 -gpuid -1
(Having put all the stuff on my Desktop for demonstration purposes)
Is this correct at all? Or am I doing something wrong there?
The error is this:
/Users/username/torch/install/bin/luajit: /Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: module 'misc.utils' not found:No LuaRocks module found for misc.utils
no field package.preload['misc.utils']
no file '/Users/username/.luarocks/share/lua/5.1/misc/utils.lua'
no file '/Users/username/.luarocks/share/lua/5.1/misc/utils/init.lua'
no file '/Users/username/torch/install/share/lua/5.1/misc/utils.lua'
no file '/Users/username/torch/install/share/lua/5.1/misc/utils/init.lua'
no file './misc/utils.lua'
no file '/Users/username/torch/install/share/luajit-2.1.0-alpha/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils/init.lua'
no file '/Users/username/.luarocks/lib/lua/5.1/misc/utils.so'
no file '/Users/username/torch/install/lib/lua/5.1/misc/utils.so'
no file './misc/utils.so'
no file '/usr/local/lib/lua/5.1/misc/utils.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
no file '/Users/username/.luarocks/lib/lua/5.1/misc.so'
no file '/Users/username/torch/install/lib/lua/5.1/misc.so'
no file './misc.so'
no file '/usr/local/lib/lua/5.1/misc.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
stack traceback:
[C]: in function 'error'
/Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
/Users/username/Desktop/neuraltalk2-master/eval.lua:7: in main chunk
[C]: in function 'dofile'
...username/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x0106ea8f10
from neuraltalk2.
Is it correct to just download the neuraltalk2-master folder, or do I need to "install" it somehow?
from neuraltalk2.
do this:
cd /Users/username/Desktop/neuraltalk2-master/
th eval.lua -model /Users/username/Desktop/checkpoint_v1.t7_cpu.t7 -image_folder /Users/username/Desktop/images -num_images 10 -gpuid -1
from neuraltalk2.
Looks like eval.lua isn't finding the local file https://github.com/karpathy/neuraltalk2/blob/master/misc/utils.lua
soumith's suggestion of cd-ing into the project directory first should work.
from neuraltalk2.
Now it complains that hdf5 is missing. Is hdf5 necessary to make it work?
/Users/username/torch/install/bin/luajit: /Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: /Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: module 'hdf5' not found:No LuaRocks module found for hdf5
no field package.preload['hdf5']
no file '/Users/username/.luarocks/share/lua/5.1/hdf5.lua'
no file '/Users/username/.luarocks/share/lua/5.1/hdf5/init.lua'
no file '/Users/username/torch/install/share/lua/5.1/hdf5.lua'
no file '/Users/username/torch/install/share/lua/5.1/hdf5/init.lua'
no file './hdf5.lua'
no file '/Users/username/torch/install/share/luajit-2.1.0-alpha/hdf5.lua'
no file '/usr/local/share/lua/5.1/hdf5.lua'
no file '/usr/local/share/lua/5.1/hdf5/init.lua'
no file '/Users/username/.luarocks/lib/lua/5.1/hdf5.so'
no file '/Users/username/torch/install/lib/lua/5.1/hdf5.so'
no file './hdf5.so'
no file '/usr/local/lib/lua/5.1/hdf5.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
stack traceback:
[C]: in function 'error'
/Users/username/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
eval.lua:8: in main chunk
[C]: in function 'dofile'
...username/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x010dd78f10
When I try to install it with these steps
brew tap homebrew/science
brew install hdf5
git clone [email protected]:deepmind/torch-hdf5.git
cd torch-hdf5
luarocks make hdf5-0-0.rockspec
Then at the step git clone [email protected]:deepmind/torch-hdf5.git
it gives me this error:
Cloning into 'torch-hdf5'...
Permission denied (publickey).
fatal: Could not read from remote repository.
Please make sure you have the correct access rights
and the repository exists.
Where does this repository need to exist?
from neuraltalk2.
Yeeeeeeeeeees!!!!!
"a man is holding a frisbee in front of a crowd"
Success, it works!
Paul, thank you SO much! Really, really appreciate your patience with a beginner like me. I got it to work at last and can now finally go to sleep happily at 5.30 am local time. Phew. Without you I would have given up long ago.
Thank you!
from neuraltalk2.
There is one litte question left (even though the issue is correctly closed already): Why do I have to reload th and luarocks via source ~/.profile
everytime after restarting the terminal? Is there a way to remember it, so I don't have to do this manually?
from neuraltalk2.
My two cents. Trying to solve dependency issues with docker: https://github.com/beeva-enriqueotero/docker-neuraltalk2
At the moment it works for CPU, not for GPU
Thanks for this awesome project!
from neuraltalk2.
Thanks @beeva-enriqueotero , I don't have too much experience with docker (still on my todo list to dive into properly). Does it work? If you could put some info into the Readme I'd be happy to link to this repo.
from neuraltalk2.
I have a similar issue on my mac running the eval lua command (th eval.lua -model .. /model_id1-501-1448236541.t7 -image_folder .. /img -num_images 1) on a small folder of ten images.
I am in the neurotalk2 folder but nonetheless I get:
/Users/localadmin/torch/install/bin/luajit: ...rs/localadmin/torch/install/share/lua/5.1/trepl/init.lua:383: module 'misc.utils' not found:No LuaRocks module found for misc.utils
no field package.preload['misc.utils']
no file '/Users/localadmin/.luarocks/share/lua/5.1/misc/utils.lua'
no file '/Users/localadmin/.luarocks/share/lua/5.1/misc/utils/init.lua'
no file '/Users/localadmin/torch/install/share/lua/5.1/misc/utils.lua'
no file '/Users/localadmin/torch/install/share/lua/5.1/misc/utils/init.lua'
no file './misc/utils.lua'
no file '/Users/localadmin/torch/install/share/luajit-2.1.0-beta1/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils.lua'
no file '/usr/local/share/lua/5.1/misc/utils/init.lua'
no file '/Users/localadmin/.luarocks/lib/lua/5.1/misc/utils.so'
no file '/Users/localadmin/torch/install/lib/lua/5.1/misc/utils.so'
no file './misc/utils.so'
no file '/usr/local/lib/lua/5.1/misc/utils.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
no file '/Users/localadmin/.luarocks/lib/lua/5.1/misc.so'
no file '/Users/localadmin/torch/install/lib/lua/5.1/misc.so'
no file './misc.so'
no file '/usr/local/lib/lua/5.1/misc.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
stack traceback:
[C]: in function 'error'
...rs/localadmin/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
eval.lua:7: in main chunk
[C]: in function 'dofile'
...dmin/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x010f243bc0
Any suggestions what I m doing wrong?
I traced back all the steps (except pointing the library path to the cudnn - I didn not get that).
from neuraltalk2.
I've just updated Readme info on https://github.com/beeva-enriqueotero/docker-neuraltalk2 with instructions to launch a docker container with neuraltalk2.
It works only for CPU mode (-gpuid -1)
from neuraltalk2.
Could you add one for gpu mode? I m trying to run the livecam version
edit: sorry saw the ToDo just now. Any suggestions on my error though? I ve now added the cudnn lib to the admin path.
from neuraltalk2.
Hi @PaulSchnau - thanks so much for assisting.
I just tried running luarocks install hdf5
, however, the error I am getting is:
Installing https://raw.githubusercontent.com/torch/rocks/master/hdf5-20-0.rockspec...
Using https://raw.githubusercontent.com/torch/rocks/master/hdf5-20-0.rockspec... switching to 'build' mode
Cloning into 'torch-hdf5'...
remote: Counting objects: 44, done.
remote: Compressing objects: 100% (36/36), done.
remote: Total 44 (delta 3), reused 23 (delta 2), pack-reused 0
Receiving objects: 100% (44/44), 28.23 KiB | 0 bytes/s, done.
Resolving deltas: 100% (3/3), done.
Checking connectivity... done.
cmake -E make_directory build;
cd build;
cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/Users/loop/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/Users/loop/torch/install/lib/luarocks/rocks/hdf5/20-0";
make
-- The C compiler identification is AppleClang 6.1.0.6020053
-- The CXX compiler identification is AppleClang 6.1.0.6020053
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Found Torch7 in /Users/loop/torch/install
CMake Error at /usr/local/Cellar/cmake/3.0.2/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:136 (message):
Could NOT find HDF5 (missing: HDF5_LIBRARIES HDF5_INCLUDE_DIRS) (Required
is at least version "1.8")
Call Stack (most recent call first):
/usr/local/Cellar/cmake/3.0.2/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:343 (_FPHSA_FAILURE_MESSAGE)
/usr/local/Cellar/cmake/3.0.2/share/cmake/Modules/FindHDF5.cmake:360 (find_package_handle_standard_args)
CMakeLists.txt:4 (FIND_PACKAGE)
-- Configuring incomplete, errors occurred!
See also "/tmp/luarocks_hdf5-20-0-7034/torch-hdf5/build/CMakeFiles/CMakeOutput.log".
make: *** No targets specified and no makefile found. Stop.
Error: Build error: Failed building.
Looks like it is failing to find HDF5
from neuraltalk2.
@PaulSchnau I just tried this - hdf5 seems to be successfully installed by Homebrew:
~/MacPorts-2.3.3 brew install homebrew/science/hdf5
==> Tapping homebrew/science
Cloning into '/usr/local/Library/Taps/homebrew/homebrew-science'...
remote: Counting objects: 571, done.
remote: Compressing objects: 100% (570/570), done.
remote: Total 571 (delta 2), reused 53 (delta 0), pack-reused 0
Receiving objects: 100% (571/571), 470.59 KiB | 0 bytes/s, done.
Resolving deltas: 100% (2/2), done.
Checking connectivity... done.
Tapped 564 formulae (592 files, 1.5M)
==> Installing hdf5 from homebrew/science
==> Installing dependencies for homebrew/science/hdf5: szip
==> Installing homebrew/science/hdf5 dependency: szip
==> Downloading https://homebrew.bintray.com/bottles/szip-2.1.yosemite.bottle.1.tar.gz
######################################################################## 100.0%
==> Pouring szip-2.1.yosemite.bottle.1.tar.gz
🍺 /usr/local/Cellar/szip/2.1: 9 files, 109.5K
==> Installing homebrew/science/hdf5
==> Downloading https://homebrew.bintray.com/bottles-science/hdf5-1.8.16_1.yosemite.bottle.tar.gz
######################################################################## 100.0%
==> Pouring hdf5-1.8.16_1.yosemite.bottle.tar.gz
🍺 /usr/local/Cellar/hdf5/1.8.16_1: 176 files, 10.5M
However, running this again:
~/neuraltalk2 ⑂ master th eval.lua -model ~/Downloads/ms_coco.t7 -image_folder ~/Desktop/fs/ -num_images 17
/Users/loop/torch/install/bin/luajit: /Users/loop/torch/install/share/lua/5.1/trepl/init.lua:383: /Users/loop/torch/install/share/lua/5.1/trepl/init.lua:383: module 'hdf5' not found:No LuaRocks module found for hdf5
no field package.preload['hdf5']
no file '/Users/loop/.luarocks/share/lua/5.1/hdf5.lua'
no file '/Users/loop/.luarocks/share/lua/5.1/hdf5/init.lua'
no file '/Users/loop/torch/install/share/lua/5.1/hdf5.lua'
no file '/Users/loop/torch/install/share/lua/5.1/hdf5/init.lua'
no file './hdf5.lua'
no file '/Users/loop/torch/install/share/luajit-2.1.0-beta1/hdf5.lua'
no file '/usr/local/share/lua/5.1/hdf5.lua'
no file '/usr/local/share/lua/5.1/hdf5/init.lua'
no file '/Users/loop/.luarocks/lib/lua/5.1/hdf5.so'
no file '/Users/loop/torch/install/lib/lua/5.1/hdf5.so'
no file '/Users/loop/torch/install/lib/hdf5.dylib'
no file './hdf5.so'
no file '/usr/local/lib/lua/5.1/hdf5.so'
no file '/usr/local/lib/lua/5.1/loadall.so'
stack traceback:
[C]: in function 'error'
/Users/loop/torch/install/share/lua/5.1/trepl/init.lua:383: in function 'require'
eval.lua:8: in main chunk
[C]: in function 'dofile'
...loop/torch/install/lib/luarocks/rocks/trepl/scm-1/bin/th:145: in main chunk
[C]: at 0x010b42ad60
from neuraltalk2.
@PaulSchnau Thank you for such a quick reply - you are awesome.
Now torch is telling me I am missing cutorch, so I typed in the following:
~/neuraltalk2 ⑂ master luarocks install cutorch
Installing https://raw.githubusercontent.com/torch/rocks/master/cutorch-scm-1.rockspec...
Using https://raw.githubusercontent.com/torch/rocks/master/cutorch-scm-1.rockspec... switching to 'build' mode
Cloning into 'cutorch'...
remote: Counting objects: 104, done.
remote: Compressing objects: 100% (98/98), done.
remote: Total 104 (delta 8), reused 42 (delta 3), pack-reused 0
Receiving objects: 100% (104/104), 146.18 KiB | 0 bytes/s, done.
Resolving deltas: 100% (8/8), done.
Checking connectivity... done.
cmake -E make_directory build && cd build && cmake .. -DCMAKE_BUILD_TYPE=Release -DCMAKE_PREFIX_PATH="/Users/loop/torch/install/bin/.." -DCMAKE_INSTALL_PREFIX="/Users/loop/torch/install/lib/luarocks/rocks/cutorch/scm-1" && make -j$(getconf _NPROCESSORS_ONLN) install
-- The C compiler identification is AppleClang 6.1.0.6020053
-- The CXX compiler identification is AppleClang 6.1.0.6020053
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Found Torch7 in /Users/loop/torch/install
CMake Error at /usr/local/Cellar/cmake/3.0.2/share/cmake/Modules/FindCUDA.cmake:568 (message):
Specify CUDA_TOOLKIT_ROOT_DIR
Call Stack (most recent call first):
CMakeLists.txt:7 (FIND_PACKAGE)
-- Configuring incomplete, errors occurred!
See also "/tmp/luarocks_cutorch-scm-1-734/cutorch/build/CMakeFiles/CMakeOutput.log".
Error: Build error: Failed building.
EDIT: Tried adding -gpuid -1
since I am running purely on CPU, however, got the following error: unknown Torch class <torch.CudaTensor>
from neuraltalk2.
I just fixed it - turns out I was using the GPU models instead of the CPU model.
from neuraltalk2.
I am trying to locate the string "Camera Dropped Frame" as I am using text-to-speech on NeuralTalk2, and I do not need the "Camera Dropped Frame" text but doing a grep of both the library on Github and on my machine, I couldn't locate this string, making me think this may be a language thing. How do I remove this string? @PaulSchnau
from neuraltalk2.
@PaulSchnau
my operating system is Mac
- brew install hdf5
- luarocks install hdf5, then i met the following error.
Could NOT find HDF5 (missing: HDF5_HL_LIBRARIES) (found suitable version
"1.8.16", minimum required is "1.8")
can someone help me? any suggestion is appreciated!
from neuraltalk2.
Hi all. Thanks for the super helpful thread. I'm encountering the same issue with luarocks install hdf5 as wzyflyleaf
(OS X 10.11.5, brew install hdf5 works, luarocks install hdf5 fails).
from neuraltalk2.
@wzyflyleaf @sebastianquack Well, finally I encounter this trouble.
After
yum install hdf5 & luarocks install hdf5
I meet
" CMake Error at /usr/share/cmake/Modules/FindPackageHandleStandardArgs.cmake:108 (message):
Could NOT find HDF5 (missing: HDF5_LIBRARIES HDF5_INCLUDE_DIRS) (Required is at least version "1.8") "
I firstly run command find / -name hdf5*
, try to find somewhere called "hdf5/lib" , but unfortunately nothing found ,
Secondly, I yum search this software, find there are
hdf5-devel.x86_64 : HDF5 development files
hdf5-mpich.x86_64 : HDF5 mpich libraries
hdf5-mpich-devel.x86_64 : HDF5 mpich development files
hdf5-mpich-static.x86_64 : HDF5 mpich static libraries
hdf5-openmpi.x86_64 : HDF5 openmpi libraries
hdf5-openmpi-devel.x86_64 : HDF5 openmpi development files
hdf5-openmpi-static.x86_64 : HDF5 openmpi static libraries
hdf5-static.x86_64 : HDF5 static libraries
h5py.x86_64 : A Python interface to the HDF5 library
hdf5.x86_64 : A general purpose library and file format for storing scientific data
I run
yum install hdf5-static
and then
'luarocks install hdf5'
BUILD PASS~
from neuraltalk2.
Related Issues (20)
- Replacing words in captions
- Missing file name HOT 1
- Few notes on the image test split of this repo.
- coco_preprocess.ipynb: print val.keys() ^ SyntaxError: invalid syntax HOT 1
- Is it possible to port this to Pytorch?
- Do you have Deep-Visual alignment cvpr 2015 implemented?
- Already captioned mscoco images
- cudnn problem HOT 3
- protobuf, loadcaffe, and hdf5 are also required for inference
- train.lua error , cuda runtime error (2) : out of memory HOT 1
- Wrong JPEG library version: library is 90, caller expects 80
- language_eval error HOT 5
- Good pretrained checkpoint model
- hi,do you think pytorch is better than torch?
- AssertionError: error: some caption had no words?
- Chinese image description, In the result, multiple words of the same type appear
- init.lua:235: Not a JPEG file: starts with 0x47 0x49 HOT 1
- train.lua run error
- how to run it in win10???
- DeepAI.org display is non-functional. Is this supposed to be the case? IDK.
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from neuraltalk2.