Error on custom layer (MATLAB EXAMPLE) (nnet.Laye​rs.Acceler​atable path cannot be found)

Hello everyone,

I literally just copied the matlab example for a custom regression layer into my file and tried to run it and I got the following error:

Error using maeRegressionLayer

The specified superclass 'nnet.layer.Acceleratable' contains a parse error, cannot be found on MATLAB's search

path, or is shadowed by another file with the same name.

If I type

which -all nnet.Layer.Acceleratable

I get:

'nnet.Layer.Acceleratable' not found.

If I remove "& nnet.layer.Acceleratable" from the code, I get the following error:

Not enough input arguments.

Error in maeRegressionLayer (line 12)

layer.Name = name;

which is really weird, because I literally just copied the example from this Matlab page: https://de.mathworks.com/help/deeplearning/ug/define-custom-regression-output-layer.html#mw_1e4dfcfe-9c30-4753-ad62-7a86a63fcfd2

[tomer roditi]

I'm having the same problem, did you find whats causing it?

[Walter Roberson]

Which MATLAB release are you using?

[ThomasP]

@tomer roditi unfortunately not :/

@Walter Roberson I'm using 21b

Answer 1

[Walter Roberson]

That layer is new as of R2022a.

[tomer roditi]

Thank you very much!