Reference to non-existent field 'gpus'.

[mjohn123]

I was successfully build matconvnet with GPU. Then I tried to run fcnTrain.m and got error as

Reference to non-existent field 'gpus'.

Error in fcnTrain (line 95)
bopts.useGpu = numel(opts.train.gpus) > 0 ;

If I comment the line bopts.useGpu = numel(opts.train.gpus) > 0 ;, I got a new error as

Error using  ' 
Transpose on ND array is not defined. Use PERMUTE instead.

Error in vl_argparse (line 76)
    values = horzcat(values, struct2cell(args{ai})') ;

Error in cnn_train_dag (line 32)
opts = vl_argparse(opts, varargin) ;

Error in fcnTrain (line 98)
info = cnn_train_dag(net, imdb, getBatchWrapper(bopts), ...

I am using MATLAB 2015b in Window OS, newest matconvnet and matconvnet-fcn versions. Someone got the same error? How can I solve it.

Thanks everybody

[islamamirul]

Hi mjohn123, I'm also getting the same error even if I enable the GPU. Did you find any way to fix it?

[mjohn123]

Hello islamamirul. I did not find any solution. I am using MATLAB 2015b. I think it maybe work in order version. But I did not try. If you find any solution. Please let me know. Thanks in adavance

[wangpichao]

I also encountered this problem. I am using Matlab 2014b/2015b/2013b under Ubuntu 14.04/windows 10. I have tried three GPUs, one K40, one GT730, one GTX 970, but on all of which the same error happens. How to handle it? Thanks in advance.

[xuranzhao711]

Hi, I encountered the same problem. If I am right, there is something wrong with the fcntrain.m script but there is an easy fix.
First search and every trainOpts and replace them by opts.train.
Then change line 23 opts.train = struct([]) to opts.train = [];
For me fixed the problem.

[mjohn123]

Thank xuranzhao711. It worked now. It is good solution

[islamamirul]

Thanks a lot xuranzhao711. It's working perfectly now.

[wangpichao]

Thanks xuranzhao711. It's working now.

[tianzq]

It works for me, thanks @xuranzhao711