Deep learning Extension (DL4J): Bug Report

varunm1

Hello,

The deep learning extension in rapidminer cannot predict label when a single sample is provided as a test file, it looks like the apply model in case of Deep learning extension needs a minimum of 2 samples. This looks like a bug.

I attached the .rmp file for process that throws a bug in this post. Below are the images related to this issue I cross-checked with@lionelderkrikor. It is evident that keras in python is able to predict on a single sample based on below images.@lionelderkrikoralso tried the keras extension in rapidminer and observed that it cannot make a prediction on a single sample as well as shown below.

DL4J extension bug:


Keras Extension in Rapidminer@lionelderkrikorproduced output.



Keras Extension in python by@lionelderkrikor:


Thanks for looking into this.@ssv112这是我们面临的问题worked on your question.

@pschlundercan you please look into this.

Thanks

pschlunder

Hi@vanumul,

thanks for pointing this out! This should not be the case. I'll look into it.