QMapDataset demo

[rscadrien]

Title: Solving the qubit mapping problem with machine learning

Summary: The demo introduces the qubit mapping problem, an NP-hard challenge in quantum computing systems with limited hardware connectivity, and explores machine learning-based approaches for solving it. It presents the QMapDataset available in PennyLane and shows how it can be used to train a Graph Neural Network to tackle the qubit mapping problem.

Relevant references:

Possible Drawbacks:

Related GitHub Issues:

---

If you are writing a demonstration, please answer these questions to facilitate the marketing process.

• GOALS — Why are we working on this now?

QMapDataset has been available in PennyLane since March 2026. The goal of this demo is to explain how to access QMapDataset in PennyLane and to show an example of how to use it for solving qubit mapping using machine learning approaches.

• AUDIENCE — Who is this for?

Quantum scientists specializing in quantum compilation and/or the use of machine learning to improve quantum computing.

• KEYWORDS — What words should be included in the marketing post?
  machine learning, quantum compilation, AI for Quantum

• Which of the following types of documentation is most similar to your file?
  (more details here)

[ ] Demo

[github-actions]

👋 Hey, looks like you've updated some demos!

🐘 Don't forget to update the dateOfLastModification in the associated metadata files so your changes are reflected in Glass Onion (search and recommendations).

Please hide this comment once the field(s) are updated. Thanks!

[drdren]

The method make_edge_mlp has not been defined. Is this how you want it to work? Please insert your actual code here.

Suggested change

	from scipy.optimize import linear_sum_assignment
	class QubitMapping_dot_Model(nn.Module):
	from scipy.optimize import linear_sum_assignment
	def make_edge_mlp(edge_dim, out_dim, hidden=32):
	return nn.Sequential(nn.Linear(edge_dim, hidden), nn.ReLU(),
	nn.Linear(hidden, out_dim))
	class QubitMapping_dot_Model(nn.Module):

[drdren]

The training and test data are not defined. You may need to adjust the format of the data.

Suggested change

	#
	def train_loop(model, optimizer, loss_fn, loader_train, loader_valid, n_epoch, device):
	"""
	#
	train_size, test_size = 0.8, 0.2
	n = len(ds.circuits)
	train_idx, val_idx = random_split(range(n), [train_size, test_size])
	def train_loop(model, optimizer, loss_fn, loader_train, loader_valid, n_epoch, device):
	"""

[drdren]

You'll need to define this before you call it.

[github-actions]

Your preview is ready 🎉!

You can view your changes here

Deployed at: 2026-07-02 18:58:44 UTC

[drdren]

This documentation might be excessive for a simple code snippet, and is incongruous with the other methods in this code block.