Clever Geek Handbook

Installing Theano, nvcc is out of the way

I installed theano on windows7,64bit on winpython using their guide http://deeplearning.net/software/theano/install_windows.html and I thought it worked since I started my first example I had the expected results and no errors. I would like to go ahead and install the part: Configure Theano to use the GPU , but when I started it again, I had this in the python console:

runfile('C:/Users/PS/Desktop/prova theano.py', wdir='C:/Users/PS/Desktop')
ERROR (theano.sandbox.cuda): nvcc compiler not found on $PATH. Check your nvcc installation and try again.
ERROR:theano.sandbox.cuda:nvcc compiler not found on $PATH. Check your nvcc installation and try again.
NP time: 0.156000[s], theano time: 0.181000[s] (times should be close when run on CPU!)
Result difference: 0.000000

... I am using the .theoror file:

[global]
device = gpu
floatX = float32

[cuda]
root = C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v5.5

[nvcc]
flags=-LC:\SciSoft\WinPython-64bit-2.7.9.4\python-2.7.9.amd64\libs
compiler_bindir=C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\bin

and I added it to settings C: \ SciSoft \ WinPython-64bit-2.7.9.4 \ as I understood from the manual.

BTW, I checked C: \ SciSoft \ env.bat, and when I write where nvcc says the file was not found, I have no problem with other checks instead. Is it because I don't have an NVIDIA card? I am completely lost. Any help? Thanks you

+3


source to share


2 answers


Theano is designed to run (almost) identically on both CPU and GPU. You don't need a GPU to use Theano, and if you don't have an Nvidia GPU you shouldn't try to install any GPU-specific stuff.



+1


source


aleju, if you don't want (or can't) use theano with a GPU, you just have to change .theoror to use only the CPU. This will not cause any problems, except for poor performance.



[global]
device = cpu
...
-1


source






More articles:


All Articles