A Python library to integrate automatic differentiation tools such as JAX with QuTiP and related quantum software packages.

This package is a work in progress. Feel free to take part in the discussions by opening new issues.

• Installation
• About
• Documentation
• Contributing
• Acknowledgements

To install qgrad development version, clone this repository and from the terminal type

python setup.py develop

qgrad dependencies are automatically installed with pip. They are:

numpy scipy matplotlib cython pytest qutip jax

qgrad is a library that implements Hamiltonian learning in the context of quantum physics-based optimization tasks. qgrad reproduces essential QuTiP functions to reduce the friction for existing QuTiP users.

qgrad leverages the powerful Python scientific stack and interfaces with the popular machine learning library JAX, to make auto-differentiation of many quantum routines possible for the desired learning tasks.

The latest documentation can be found here. It includes the API reference and examples.

We are in the early stages of designing the tool and welcome any discussion in the form of Issues or Pull Requests.

This package started as part of @araza6's GSoC 2020 project. All the work relevant to GSoC 2020 is compiled in this release: https://github.com/qgrad/qgrad/releases/tag/0.0.dev2