You are using the old Theano gpu back-end. Its installation on Windows was very complicated. I strongly recommand to use the new gpu back-end. Its installation on windows is much simpler. We also don't have time to help with the old back-end anymore. See this doc to use the new back-end:
https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29 Fred On Thu, Apr 13, 2017 at 12:50 AM davidstarxxxxx <[email protected]> wrote: > > down votefavorite > <http://stackoverflow.com/questions/43371190/cant-use-gpu-with-theano-win10#> > > I builded and executed deviceQuery.exe with VS2015, and the result was > Pass, but when I test theano with the code below, some errors occurred. > > from theano import function, config, shared, tensorimport numpyimport time > > vlen = 10 * 30 * 768 # 10 x #cores x # threads per core > iters = 1000 > > rng = numpy.random.RandomState(22) > x = shared(numpy.asarray(rng.rand(vlen), config.floatX)) > f = function([], tensor.exp(x))print(f.maker.fgraph.toposort()) > t0 = time.time()for i in range(iters): > r = f() > t1 = time.time()print("Looping %d times took %f seconds" % (iters, t1 - > t0))print("Result is %s" % (r,))if numpy.any([isinstance(x.op, > tensor.Elemwise) and > ('Gpu' not in type(x.op).__name__) > for x in f.maker.fgraph.toposort()]): > print('Used the cpu')else: > print('Used the gpu') > > Error: > > c1xx: fatal error C1083: Cannot open source file: 'mod.cu': No such file or > directory > nvcc warning : The 'compute_20', 'sm_20', and 'sm_21' architectures are > deprecated, and may be removed in a future release (Use > -Wno-deprecated-gpu-targets to suppress warning). > mod.cu > ['nvcc', '-shared', '-O3', '-LC:\\Users\\asus\\Anaconda3\\envs\\DL\\libs', > '-use_fast_math', '--compiler-bindir', 'C:\\Program Files (x86)\\Microsoft > Visual Studio 14.0\\VC\\bin\\x86_amd64', '-Xlinker', '/DEBUG', '-D > HAVE_ROUND', '-m64', '-Xcompiler', > '-DCUDA_NDARRAY_CUH=mc72d035fdf91890f3b36710688069b2e,-DNPY_NO_DEPRECATED_API=NPY_1_7_API_VERSION,/Zi,/MD', > > '-I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\theano\\sandbox\\cuda"', > > '-I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\numpy\\core\\include"', > '-I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\include"', > '-I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\theano\\gof"', > '-L"C:\\Users\\asus\\Anaconda3\\envs\\DL\\libs"', > '-L"C:\\Users\\asus\\Anaconda3\\envs\\DL"', '-o', > 'C:\\Users\\asus\\AppData\\Local\\Theano\\compiledir_Windows-10-10.0.14393-SP0-Intel64_Family_6_Model_94_Stepping_3_GenuineIntel-3.5.3-64\\cuda_ndarray\\cuda_ndarray.pyd', > 'mod.cu', '-lcublas', '-lpython35', '-lcudart'] > ERROR (theano.sandbox.cuda): Failed to compile cuda_ndarray.cu: ('nvcc return > status', 2, 'for cmd', 'nvcc -shared -O3 > -LC:\\Users\\asus\\Anaconda3\\envs\\DL\\libs -use_fast_math --compiler-bindir > C:\\Program Files (x86)\\Microsoft Visual Studio 14.0\\VC\\bin\\x86_amd64 > -Xlinker /DEBUG -D HAVE_ROUND -m64 -Xcompiler > -DCUDA_NDARRAY_CUH=mc72d035fdf91890f3b36710688069b2e,-DNPY_NO_DEPRECATED_API=NPY_1_7_API_VERSION,/Zi,/MD > > -I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\theano\\sandbox\\cuda" > > -I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\numpy\\core\\include" > -I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\include" > -I"C:\\Users\\asus\\Anaconda3\\envs\\DL\\lib\\site-packages\\theano\\gof" > -L"C:\\Users\\asus\\Anaconda3\\envs\\DL\\libs" > -L"C:\\Users\\asus\\Anaconda3\\envs\\DL" -o > C:\\Users\\asus\\AppData\\Local\\Theano\\compiledir_Windows-10-10.0.14393-SP0-Intel64_Family_6_Model_94_Stepping_3_GenuineIntel-3.5.3-64\\cuda_ndarray\\cuda_ndarray.pyd > mod.cu -lcublas -lpython35 -lcudart') > WARNING (theano.sandbox.cuda): The cuda backend is deprecated and will be > removed in the next release (v0.10). Please switch to the gpuarray backend. > You can get more information about how to switch at this URL: > > https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end%28gpuarray%29 > > WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not > available (error: cuda unavailable) > > It seems that cuda_ndarray.cu can't be compiled, but I can't figure out > the reason. > > My .theanorc file > > [global] > > openmp=False > > device = gpu > > optimizer_including=cudnn > > floatX = float32 > > allow_input_downcast=True > [lib] > > cnmem = 0.8 > [blas] > > ldflags= > [gcc] > > cxxflags=-IC:\Users\asus\Anaconda3\envs\DL\MinGW > [nvcc] > > flags = -LC:\Users\asus\Anaconda3\envs\DL\libs > > compiler_bindir = C:\Program Files (x86)\Microsoft Visual Studio > 14.0\VC\bin\x86_amd64 > > fastmath = True > > cuda version : 8.0 > > cuDNN version : 5.1 > > theano version : 0.9.0 (conda install theano) > > visual studio version : 2015 > > operating system : win10 (64-bit) > > python version : 3.5.3 (64-bit) > > I have spent over 15 hours googling and configuring it, and I am willing > to add any essential information. I appreciate for any advice. > > -- > > --- > You received this message because you are subscribed to the Google Groups > "theano-users" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. > -- --- You received this message because you are subscribed to the Google Groups "theano-users" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
