I have been having the following problem for a few days and can't seem to get around it.
I have installed Theano on Ubuntu using the link provided by Lasagne: pip install -r https://raw.githubusercontent.com/Lasagne/Lasagne/v0.1/requirements.txt When I test Theano from the command line using the following: THEANO_FLAGS=device=gpu python -c "import theano; print(theano.sandbox.cuda.device_properties(0))" The output is: Using gpu device 0: GeForce GTX 980 (CNMeM is disabled, cuDNN 5005) {'major': 5, 'tccDriver': 0, 'kernelExecTimeoutEnabled': 1, 'deviceOverlap': 1, 'driverVersion': 8000, 'warpSize': 32, 'concurrentKernels': 1, 'maxThreadsPerBlock': 1024, 'computeMode': 0, 'canMapHostMemory': 1, 'maxGridSize2': 65535, 'maxGridSize1': 65535, 'maxGridSize0': 2147483647, 'integrated': 0, 'minor': 2, 'ECCEnabled': 0, 'runtimeVersion': 7050, 'textureAlignment': 512, 'multiProcessorCount': 16, 'clockRate': 1278500, 'totalConstMem': 65536, 'name': 'GeForce GTX 980', 'memPitch': 2147483647, 'maxThreadsDim1': 1024, 'maxThreadsDim0': 1024, 'maxThreadsDim2': 64, 'coresCount': -16, 'sharedMemPerBlock': 49152, 'regsPerBlock': 65536} So, everything seems to be working fine. However, when I open IPython, either from the terminal or using and IPython notebook and I try to import Theano, I get an enourmous printout of 5327 lines, of what I presume is the definition of CUDA_NDARRAY. This is followed by many warnings which I will leave out for now, but am happy to post if requested. The printout ends with the following Error and Warning: ['nvcc', '-shared', '-O3', '-m64', '-Xcompiler', '-DCUDA_NDARRAY_CUH=mc72d035fdf91890f3b36710688069b2e,-DNPY_NO_DEPRECATED _API=NPY_1_7_API_VERSION,-fPIC,-fvisibility=hidden', '-Xlinker', '-rpath,/home/yossi/.theano/compiledir_Linux-4.4--generic -x86_64-with-Ubuntu-16.04-xenial-x86_64-3.5.2-64/cuda_ndarray', '-I/usr/local/lib/python3.5/dist-packages/theano/sandbox/c uda', '-I/usr/local/lib/python3.5/dist-packages/numpy/core/include', '-I/usr/include/python3.5m', '-I/usr/local/lib/python 3.5/dist-packages/theano/gof', '-o', '/home/yossi/.theano/compiledir_Linux-4.4--generic-x86_64-with-Ubuntu-16.04-xenial-x8 6_64-3.5.2-64/cuda_ndarray/cuda_ndarray.so', 'mod.cu', '-L/usr/lib', '-lpython3.5m', '-lcublas', '-lcudart'] ERROR (theano.sandbox.cuda): Failed to compile cuda_ndarray.cu: ('nvcc return status', 1, 'for cmd', 'nvcc -shared -O3 -m6 4 -Xcompiler -DCUDA_NDARRAY_CUH=mc72d035fdf91890f3b36710688069b2e,-DNPY_NO_DEPRECATED_API=NPY_1_7_API_VERSION,-fPIC,-fvisi bility=hidden -Xlinker -rpath,/home/yossi/.theano/compiledir_Linux-4.4--generic-x86_64-with-Ubuntu-16.04-xenial-x86_64-3.5 .2-64/cuda_ndarray -I/usr/local/lib/python3.5/dist-packages/theano/sandbox/cuda -I/usr/local/lib/python3.5/dist-packages/n umpy/core/include -I/usr/include/python3.5m -I/usr/local/lib/python3.5/dist-packages/theano/gof -o /home/yossi/.theano/com piledir_Linux-4.4--generic-x86_64-with-Ubuntu-16.04-xenial-x86_64-3.5.2-64/cuda_ndarray/cuda_ndarray.so mod.cu -L/usr/lib -lpython3.5m -lcublas -lcudart') WARNING (theano.sandbox.cuda): CUDA is installed, but device gpu is not available (error: cuda unavilable) I'm really not sure what this means or how to fix it. I have tried uninstalling and reinstalling Theano (including using different versions). I have included the following to my .theanorc: [global] floatX = float32 device = gpu I've also added the following to my .bashrc: export CUDA_ROOT=/usr/local/cuda-7.5 export PATH=$PATH:$CUDA_ROOT/bin export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$CUDA_ROOT/lib64 export THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 I'm not sure where to go from here. Any advice would be much appreciated! -- --- 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 theano-users+unsubscr...@googlegroups.com. For more options, visit https://groups.google.com/d/optout.
