ERROR cudaGetDeviceCount

[gabriellimagomes15]

Hello,

I am running Somoclu in GPU, but I got this error

RuntimeError Traceback (most recent call last)
in ()
2 n_rows, n_columns = 10,10
3 som = somoclu.Somoclu(n_columns, n_rows, kerneltype=1)
----> 4 som.train(data, epochs=10)
5 #som.view_umatrix(bestmatches=True, bestmatchcolors=colors, labels=labels)

~/somoclu/src/Python/somoclu/train.py in train(self, data, epochs, radius0, radiusN, radiuscooling, scale0, scaleN, scalecooling)
226 self._compact_support, self._neighborhood == "gaussian",
227 self._std_coeff, self._verbose, self.codebook, self.bmus,
--> 228 self.umatrix)
229 self.umatrix.shape = (self._n_rows, self._n_columns)
230 self.bmus.shape = (self.n_vectors, 2)

RuntimeError: CUDA error calling "cudaGetDeviceCount(&devCount)", code is 35

Does Somebody can help, please?

[xgdgsc]

This is usually caused by a GPU that is turned off

[gabriellimagomes15]

Very strange because my GPU is ok.

nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver Copyright (c) 2005-2013 NVIDIA Corporation Built on Wed_Jul_17_18:36:13_PDT_2013 Cuda compilation tools, release 5.5, V5.5.0

[xgdgsc]

Seems code 35 is CUDA driver version is insufficient for CUDA runtime version https://stackoverflow.com/questions/41409842/ubuntu-16-04-cuda-8-cuda-driver-version-is-insufficient-for-cuda-runtime-vers

[gabriellimagomes15]

oooohhh my God!!!!
Thanks very much!!

[gabriellimagomes15]

@xgdgsc , I have another question.

Why the get_bmus() return just until 9 class? Do you know?

[xgdgsc]

I don' t know.

[gabriellimagomes15]

Thanks very much!