Error in All Cats & Dogs

I downloaded Roberto Cadili’s workflow “All Cats & Dogs? Classify Images with Codeless CNNs” at All Cats & Dogs? Classify Images with Codeless CNNs | KNIME. I ran it exactly as it was downloaded, of course adding the images to the folder where it was looking for them.

All the nodes in Step 1 Define Network Architecture have green dots and seem to have worked fine. Similarly, all the nodes in Step 2 Read and Preprocess Images have green dots and seem to have worked fine, as does the Conda Environment propagation.

But the processing stops at the Keras Network Learner. I get this error:

“Invalid settings: Selected input columns do not provide enough elements (1) to populate all neurons (67500) of network input ‘input_1_0:0’. Try adding some columns to the selection.”

Can someone provide suggestions as to what I should be doing at this point? Thank you!

For another reason, I installed several other extensions that KNIME didn’t tell me I needed. But now Keras Network Learner works. I wish I knew which extension was needed for that to happen.

@njacknis I once edited a cats and dogs sample workflow and also wrote down which extensions were needed. Maybe you can check that out.

There are also environment propagations for Windows and MacOS (Intel):

Thank you very much! This is very helpful.

@njacknis if you want to explore the setup of KNIME and Deep Learning further you can read this article.

This topic was automatically closed 90 days after the last reply. New replies are no longer allowed.