Quantum-Train LSTM Demo

[Abeeekoala]

General information

Name Chu-Hsuan Abraham Lin

Affiliation (optional) Imperial College London

---

Demo information

Title Quantum-Train Long Short-Term Memory (QT-LSTM) for Wave Prediction (PennyLane Public Demo)

Abstract This tutorial features the Quantum Train architecture proposed by Chen-Yu Liu et al. You can find more details on arXiv:2405.11304 and arXiv:2402.16465. Originally, in Deloitte’s Quantum Climate Challenge 2024, we implemented QT-LSTM to predict flood events along the Wupper River. However, due to the confidentiality of the data, we can only demonstrate QT-LSTM with fabricated data for simplicity while showcasing the power of this framework. In this tutorial, we implemented a classical LSTM model to learn the time-series wave data. We then explore the components of the LSTM model and finally implement the QT-LSTM.

Relevant links GitHub Repo

[CatalinaAlbornoz]

Hi @Abeeekoala , thank you for submitting this demo! 🙌 We'll review it and get back to you in the next few days.

[ikurecic]

Hi @Abeeekoala , thanks so much for sharing the code from your project, and congrats again. ☺

There are a couple of things that might be nice to add, if you could take a quick look at them here:

• In the torch.save(model_qt, 'model_/model_loss_bsf') line it would be worth it to add a comment to note that it's a local folder.
• Could you make a comment in the repository readme or in the requirements file about the package versions you used to run the code? This would help with reproducibility.
• Can you confirm that the authors for this code project should be Chen-Yu Liu and yourself, like listed in the repository?
• It would be great if you could add the titles of the papers in the repository readme together with the arXiv links and designations that you already have.
• I've noticed a tiny typo, 'pakages' instead of 'packages' right at the start of the notebook.
• And, finally, would you want our Marketing Team to tag you on social media? ☺ If you share your (and/or Chen-Yu Liu's) LinkedIn and/or Twitter/X profiles, I will be able to forward those to them.

This looks great, I'm looking forward to sharing it!

[Abeeekoala]

Hi @ikurecic, thanks for the feedback we have updated the repository. In response to your questions

• Yes, Chen-Yu Liu and I are the authors of this code project.
• Yes, we are happy to be tagged. Our LinkedIn profiles are Chu-Hsuan Abraham Lin and Chen-Yu Liu

Thanks again for your invitation to this demo and we look forward to hearing from you soon.

[ikurecic]

Hi @Abeeekoala , that's brilliant, thanks so much! ☺ It should be live here either later today or on Monday, and our Marketing Team will tag you both on LinkedIn in the next couple of weeks.

Great job and congrats again on your hard work, this is really cool and I'm looking forward to seeing where you folks go next. :)

[Abeeekoala]

That is fantastic, thank you so much.

Ivana Kurečić @.***> 於 2024年8月3日 週六 上午12:26寫道：

Hi @Abeeekoala https://github.com/Abeeekoala , that's brilliant, thanks so much! ☺ It should be live here https://pennylane.ai/qml/demos_community/ either later today or on Monday, and our Marketing Team will tag you both on LinkedIn in the next couple of weeks.

Great job and congrats again on your hard work, this is really cool and I'm looking forward to seeing where you folks go next. :)

— Reply to this email directly, view it on GitHub https://github.com/PennyLaneAI/qml/issues/1175#issuecomment-2265747900, or unsubscribe https://github.com/notifications/unsubscribe-auth/A355S2WIKABZYSSBJXJXOYTZPOXMJAVCNFSM6AAAAABLVP4CQKVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDENRVG42DOOJQGA . You are receiving this because you were mentioned.Message ID: @.***>