1

I am on a Linux 3.16.0-30-generic #40~14.04.1-Ubuntu x86_64 GNU/Linux machine with NVIDIA Corporation GF108 [GeForce GT 430] [10de:0de1] (rev a1) graphic card.

I am trying to run lasagne package with GPU enabled, however by running import lasagne.layers.dnn, I get an ImportError:

Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/home/ali/DiabeticRethinopathy/diabeticrethinopathy/dr2/src/lasagne-master/lasagne/layers/dnn.py", line 12, in <module>
    raise ImportError("dnn not available")  # pragma: no cover
ImportError: dnn not available

Currently, I have installed and successfully compiled all the cuda_7.0.28_linux samples as well as cudnn-7.0-linux-x64-v3.0-rc.

python -c 'from theano.sandbox.cuda.dnn import dnn_available; print dnn_available(); print dnn_available.msg'

I get this output:

False
Device not supported by cuDNN

I am using python 2.7 and used pip to install all these packages from their tar.gz files, which are downloaded from their repositories.

P. Camilleri
  • 12,664
  • 7
  • 41
  • 76
Ali Shakiba
  • 1,255
  • 2
  • 15
  • 29

1 Answers1

1

I finally figured out the problem! My system's GPU cannot/is not supported by cudNN, unfortunately. :(

Ali Shakiba
  • 1,255
  • 2
  • 15
  • 29