Local Unbuntu 16.04 and install-gpu.sh fails in notebook

(David Clement) #1

Hi all,
After a clean install of Ubuntu and running install-gpu.sh from https://github.com/fastai/courses/blob/master/setup/install-gpu.sh seems to run fine. The machine is a high spec intel box with a Titan V. I followed all the instructions including downloading the data etc.
When I run jupyter notebook and open https://github.com/fastai/courses/blob/master/deeplearning1/nbs/lesson1.ipynb I get to cell with the code

import utils; reload(utils)
from utils import plots

and it gives me a long error ending in:

ValueError: You are tring to use the old GPU back-end. It was removed from Theano. Use device=cuda* now. See https://github.com/Theano/Theano/wiki/Converting-to-the-new-gpu-back-end(gpuarray) for more information

I have seen posts on building up combinations of libraries on a local machine however I would prefer to just use the standard installer.

I little guidance as to next steps would be much appreciated.