Thank you. It seems that Cholesky uses the GPU only for computing the gradient. Am I correct?
On Thursday, November 10, 2016 at 5:00:51 PM UTC+1, nouiz wrote: > > I answered on stackoverflow. > > On Tue, Nov 8, 2016 at 11:17 AM, Kiuhnm Mnhuik <[email protected] > <javascript:>> wrote: > >> You can find my question also on stackoverflow: >> http://stackoverflow.com/q/40487105/4603642 >> >> >> I built a neural network which needs to use the Cholesky decomposition >> and solve triangular systems as part of its computation. This means that I >> also need to compute the gradients of the whole computation, of course. >> >> When I try to compile the code I get the error "No Cula available". >> Unfortunately, I can't download Cula from the >> http://www.culatools.com/downloads/dense/ >> >> I was wondering if it's possible to mix CPU and GPU with theano. The >> matrices I need to use Cholesky and Solve on are small (100x100) so I could >> do that part on the CPU. To do so, though, I'd need to transfer the >> matrices to the CPU right after they've been computed by the GPU and then >> send the result back to the GPU. Is it possible to do this transparently >> enough? >> >> -- >> >> --- >> 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] <javascript:>. >> 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.
