Hi, The first implementation didn't supported that. Instead of just adding that change to the current GPU back-end, we desided to reimplement a new one with other important change. You can use the new gpu back-end, it support int:
http://deeplearning.net/software/theano/tutorial/using_gpu.html#gpuarray Use the development version of Theano. It received much change since the last release: http://www.deeplearning.net/software/theano/install.html#bleeding-edge-install-instructions Frédéric On Mon, Jul 18, 2016 at 11:12 PM, Shang Lee <[email protected]> wrote: > I search for that but found nothing. I try to read the code but found > that's not an easy thing. > Does any one know that or any article about that? I'm trying to write some > custom sparse Op for my task, it need use int. > > Any help appreciated. > > -- > > --- > 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. > -- --- 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.
