Daniel Galvez
Daniel Galvez
Thrust is the best bet for a device_vector container: https://github.com/NVIDIA/thrust
Hi @csukuangfj unfortunately, std::vector's operator[] (as well as at()) are not marked as __device__ functions, so cuda kernels cannot call the std::vector methods. You could use std::vector as a container...
It looks like you're running nnet-component-test. Ah I believe it is segfaulting because it is trying to run in CPU mode, not GPU mode. I hope I mentioned this earlier...
By the way, I tested this on some GeForce GPU not on the CLSP cluster; but I doubt that is the cause of the issue.
Yes. But to be sure, here is how I define "x-stride": the number of elements moved along the x dimension after each convolution is performed, in preparation for the next...
Busy finishing up an exam, but I will mention that CUDNN v5 requires CUDA 7.5. I think CLSP has 7.0 On Thu, May 19, 2016 at 2:18 PM, Daniel Povey...
@tomkocse Hi Tom, I'm guessing that you're asking this because you are unable to get similar training convergence between the CUDNN3DConvolution component and the existing Convolution component. By the way,...
Alright, finally back on this. Sorry for delay. I'll start by considering the rest of @freewym's comments.
@tomkocse Can you try the most recent code? I just pushed a series of commits. Again, let me know if there are problems. I have not added a test that...
@tomkocse I mean, in your personal repository ([this page](https://github.com/tomkocse/kaldi)), what branch are you using? I am curious what changes you made to make the 3d -> 5d fix. It is...