8
$\begingroup$

Bug introduced in 11.0.0 and fixed via paclet update


The Wolfram Mathematica site provides this example.

I have copied and pasted into a notebook. The resource is reached and correctly sampled.

However, the following does not run.

convnet = 
 NetChain[{ConvolutionLayer[20, {5, 5}], ElementwiseLayer[Ramp], 
   PoolingLayer[{2, 2}, {2, 2}], ConvolutionLayer[50, {5, 5}], 
   ElementwiseLayer[Ramp], PoolingLayer[{2, 2}, {2, 2}], 
   FlattenLayer[], DotPlusLayer[500], ElementwiseLayer[Ramp]}, 
  "Input" -> NetEncoder[{"Image", {32, 32}}]]

The following error is produced:

enter image description here

Apologies for the size.

Have other users experience the same error?

(I am on Windows 10).

$\endgroup$
  • $\begingroup$ I have contacted Technical Support. Will update if and when a reply. $\endgroup$ – ubpdqn Aug 19 '16 at 0:36
  • $\begingroup$ I have the same error - Windows 7 $\endgroup$ – andre314 Aug 19 '16 at 0:54
  • 1
    $\begingroup$ It runs fine - Mma 11.0.0 Win7 $\endgroup$ – JungHwan Min Aug 19 '16 at 0:54
  • $\begingroup$ Same error. (Using version 11.0.0 under Win 10, too.) $\endgroup$ – Karsten 7. Aug 19 '16 at 0:59
  • $\begingroup$ Same problem with this example. I have not tried other ones. $\endgroup$ – andre314 Aug 19 '16 at 1:01
11
$\begingroup$

This is actually a bug in a separate paclet that corrupts kernel state. It happens to show up here in neural networks. It has been fixed and a patch issued, your local install should acquire the update in time, but you should be able to force an update by running PacletUpdate["Streaming"].

In the meantime (or if that doesn't work for you), you can avoid this issue by running NetChain; on a fresh kernel (i.e. Quit) before running the example code.

$\endgroup$
  • $\begingroup$ thank you very much for answer. $\endgroup$ – ubpdqn Aug 24 '16 at 20:39
  • $\begingroup$ Did this issue first appear in 11.0.0? I wish to add an appropriate bugs header, per site convention. $\endgroup$ – Mr.Wizard Aug 31 '16 at 7:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.