Quantum Machine Learning Toolbox


The Quantum Machine Learning Toolbox (QMLT) is a Strawberry Fields application that simplifies the optimization of variational quantum circuits (also known as parametrized quantum circuits). Tasks for the QMLT range from variational eigensolvers and unitary learning to supervised and unsupervised machine learning with models based on a parametrized circuit.


The idea of variational quantum circuits is to classically optimize the gate parameters according to an objective.

The quantum machine learning toolbox is designed to be used on top of the Strawberry Fields’ continuous-variable quantum computing simulator, but can in principle be combined with other quantum programming languages as well.


The Quantum Machine Learning Toolbox supports:

  • The training of user-provided variational circuits
  • Automatic and numerical differentiation methods to compute gradients of circuit outputs
  • Optimization, supervised and unsupervised learning tasks
  • Regularization of circuit parameters
  • Logging of training results
  • Monitoring and visualization of training through matplotlib and TensorBoard
  • Saving and restoring trained models
  • Parallel computation/GPU usage for TensorFlow-based models

Getting started

To install the QMLT on your system, begin at the download and installation guide. Then, familiarise yourself with variational circuits and things you can do with them.

For getting started with writing your QMLT code, have a look at the tutorials for the numerical learner and the TensorFlow learner.

Finally, detailed documentation on the QMLT code is provided.


If you are having issues, please let us know, either by email or by posting the issue on our GitHub issue tracker.

We have a mailing list located at: support@xanadu.ai.


The Quantum Machine Learning Toolbox is free and open-source, released under the Apache License, Version 2.0.

Indices and tables