search

Rlag1998 / QML_Generalization

Generalization Performance of Quantum Metric Learning Classifiers
Apache License 2.0
3 stars 0 forks source link

pennylane demo integration #1

Open KetpuntoG opened 1 year ago

KetpuntoG commented 1 year ago

Hey! It's Guillermo, I will be in charge of the integration and review process of your demo. Thanks for waiting 🙂

I was wondering if you could do a PR to the qml repo with the demo integration. This would facilitate the review process. In the readme of the repo there are some instructions on how to do this. It shouldn't take long since you have it done in sphinx. If you have any questions feel free to write me 🚀

Rlag1998 commented 1 year ago

Hey Guillermo,

Thanks for getting back to us!

I have now sent a PR to the qml repo (#606) 😃

Rlag1998 commented 1 year ago

There's a couple things I feel I should point out about the demo:

  1. In the PR, there are three proposed demos. The main, most 'important' one is the one titled tutorial_embedding_generalization and it demonstrates desired generalization outcomes. The other two are sort of supplementary, as they illustrate poor generalization performance. The contents of each of the three demos are otherwise all quite similar.
  2. The code only works on versions of pennylane 0.18.0 and prior. I think it could have something to do with the mixed classical/quantum training paramaters not all being recognised as such, but I'm not 100% sure. The code works perfectly fine on version 0.18.0 though.

Thanks!

KetpuntoG commented 1 year ago

Hey @Rlag1998 ! Thanks for open the issue to the community demo integration! In the case of directing to github it would be much more convenient to have a notebook instead of the .py format of sphinx. This way, it would render directly on the web and everyone could see it 😊 Could you make the change?

Rlag1998 commented 1 year ago

Hi @KetpuntoG, I've now put all three demos in Jupyter Notebook format. I've put them in a new repository. 😄

Rlag1998 commented 1 year ago

I will also update the issue to include the new repository link

KetpuntoG commented 1 year ago

I just saw it! Great, this is just what I was looking for! a good README and the 3 notebooks. Now it's my turn to work on it :) I'll do some final checks and get it ready to be on the web. thank you @Rlag1998