Neeraj N Sajjan

Results 3 comments of Neeraj N Sajjan

Barring the fact that it will be terribly slow, training on cpu is possible. On Tue, May 8, 2018 at 3:35 PM, Ammar Qammaz wrote: > Intel GPU does not...

I had tried that before, does not seem to be affecting the output at all.

To apply the same convolution to each of the time steps independently