cudnn_conv_layer.cpp:53] Check failed: status == CUDNN_STATUS_SUCCESS (4 vs. 0) CUDNN_STATUS_INTERN

[gattupall...@gmail.com]

Hi,

I am trying to run Caffe on Jetson TX2. I get this error cudnn_conv_layer.cpp:53] Check failed: status == CUDNN_STATUS_SUCCESS (4 vs. 0)  CUDNN_STATUS_INTERNAL_ERROR.

I have compiled caffe with cuda 8.0 and cudnn 5.1. Let me know if anyone has an idea why is it failing.

Thanks!

[gattupall...@gmail.com]

It does the same thing for CPU or GPU mode. Pls Help!

[gattupall...@gmail.com]

Ok, Updates: In case anyone faces this issue. 

1.) Check permissions. Sometimes it runs by adding sudo.

2.) It is actually out of memory. I was running ResNet101 so it did this. It ran AlexNet correctly.

[gattupall...@gmail.com]

On Wednesday, May 10, 2017 at 5:24:42 PM UTC-7, gattupall...@gmail.com wrote:

Ok, its is Solved!! Updates: In case anyone faces this issue. 

[Zi Menglan]

Hi gattupall,

I have met this problem right ( I run resnet101 on tx2), how did you solve this problem?

Thanks.

[zouzh...@gmail.com]

Hi!

    Maybe you could use command "su" to get root right, and the problem will be solved!

在 2017年6月21日星期三 UTC+8上午10:54:04，Zi Menglan写道：