I am facing exactly the same issue on a cluster. Anyone has an idea how to fix it? Regards,
Le vendredi 22 janvier 2016 13:08:22 UTC-8, Tejas Khot a écrit : > > I have a fresh nvidia driver-361.18, cuda-7 and cudnn-6.5 installation > done on my Ubuntu 14.04 machine. > My gpu is detected and I can run tensorflow smoothly. However, I get a > segmentation fault on importing theano. > > theano-cache clear > doesn't seem to solve it. > > Here is the output: > Python 2.7.6 (default, Jun 22 2015, 17:58:13) > Type "copyright", "credits" or "license" for more information. > > IPython 4.0.2 -- An enhanced Interactive Python. > ? -> Introduction and overview of IPython's features. > %quickref -> Quick reference. > help -> Python's own help system. > object? -> Details about 'object', use 'object??' for extra details. > > In [1]: import tensorflow > I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA > library libcublas.so.7.0 locally > I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA > library libcudnn.so.6.5 locally > I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA > library libcufft.so.7.0 locally > I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA > library libcuda.so locally > I tensorflow/stream_executor/dso_loader.cc:101] successfully opened CUDA > library libcurand.so.7.0 locally > > In [2]: import theano > Segmentation fault > > > I have tried reinstalling theano and upgrading as well using > sudo pip install Theano --upgrade > but this error still persists. > > Contents of .theanorc > [global] > floatX=float32 > device=gpu0 > > [blas] > ldflags = -L/usr/local/lib -lopenblas > > [nvcc] > fastmath=True > > [cuda] > root=/usr/local/cuda > > I haven't faced this issue with theano in the past. Can someone help me > solve this please. > > -- --- 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.
