Hi @sophchoe, here’s the doc entry for Keras Layer, and here you can find the source code. It’s not a plugin though.
Your CV Classifier looks very interesting. Be sure to submit it as a demo here when you’re ready!
Hi @sophchoe,
I wrote to you on GitHub regarding your latest demo. Would you like to review the comments and respond on the issue? This way we can publish your demo on our website.
The correction to the abstract looks good.
For the code, I can change init_layer(inputs) to encode_data(inputs) and reupload.
I learned that the qnn keras layer does the automatic initialization of the quantum circuit parameters, so that can be simplified.
Additionally, can the circuits be simplified using “for” loops? Then the codes need not be all long and clunky.
Please let me know.
Hi @sophchoe, yes, you can simplify using the updates you mention and let me know when you’re done in order to review and publish the demo!
Hi @CatalinaAlbornoz I am working on it.
Another question to you. Is it possible to use 26*26 = 676 simulation devices concurrently? It is for a quantum convolutional network idea I have. Thank you.
Hi @sophchoe, do you mean concurrent modes as defined here? Or do you mean in parallel? If you mean in parallel then several devices allow you to use batches, which execute multiple circuits in parallel. PennyLane also allows you to execute qnodes in parallel so you could use the pennylane-sf plugin too.
