Xanadu Announces PennyLane Machine Learning Software for Quantum Computers

Xanadu calls their PennyLane software the first dedicated machine learning software for quantum computers and say they envision it as the TensorFlow for quantum computing.  The software is open source and hardware agnostic with plug-ins that can support different backends.   Currently they support integration with their own Strawberry Fields software for their continuously variable photonic quantum architecture and also Project Q from ETH Zurich for qubit based machines.  Since Project Q supports the IBM Q processors, programs developed with PennyLane can run on the IBM Q network.

PennyLane’s core feature is its ability to compute gradients of variational quantum circuits in a way compatible with classical techniques such as backpropagation. PennyLane extends the automatic differentiation algorithms common in optimization and machine learning to include quantum and hybrid computations. PennyLane can be used for the optimization of variational quantum eigensolvers (VQE), quantum approximate optimization (QAOA), quantum machine learning (QML) models, and other applications.

For more details, you can view Xanadu’s press announcement here, an arXiv paper authored by Xanadu personnel here, the PennyLane documentation here, and the PennyLane software on GitHub here.