PennyLane：混合量子-经典计算自动微分
量子物理
2022-08-02 v4 新兴技术
机器学习
计算物理
摘要
PennyLane 是一个用于量子计算机可微编程的 Python 3 软件框架。该库为近期量子计算设备提供了统一架构，支持量子比特和连续变量两种范式。PennyLane 的核心特性是能够以兼容反向传播等经典技术的方式计算变分量子电路的梯度。因此，PennyLane 将优化与机器学习中常见的自动微分算法扩展至包含量子与混合计算。插件系统使该框架兼容任意基于门的量子模拟器或硬件。我们为包括 Xanadu Cloud、Amazon Braket 和 IBM Quantum 在内的硬件提供商提供插件，允许在公开可访问的量子设备上运行 PennyLane 优化。在经典方面，PennyLane 与 TensorFlow、PyTorch、JAX 和 Autograd 等加速机器学习库接口。PennyLane 可用于变分量子本征求解器、量子近似优化、量子机器学习模型及许多其他应用的优化。
引用
@article{arxiv.1811.04968,
title = {PennyLane: Automatic differentiation of hybrid quantum-classical computations},
author = {Ville Bergholm and Josh Izaac and Maria Schuld and Christian Gogolin and Shahnawaz Ahmed and Vishnu Ajith and M. Sohaib Alam and Guillermo Alonso-Linaje and B. AkashNarayanan and Ali Asadi and Juan Miguel Arrazola and Utkarsh Azad and Sam Banning and Carsten Blank and Thomas R Bromley and Benjamin A. Cordier and Jack Ceroni and Alain Delgado and Olivia Di Matteo and Amintor Dusko and Tanya Garg and Diego Guala and Anthony Hayes and Ryan Hill and Aroosa Ijaz and Theodor Isacsson and David Ittah and Soran Jahangiri and Prateek Jain and Edward Jiang and Ankit Khandelwal and Korbinian Kottmann and Robert A. Lang and Christina Lee and Thomas Loke and Angus Lowe and Keri McKiernan and Johannes Jakob Meyer and J. A. Montañez-Barrera and Romain Moyard and Zeyue Niu and Lee James O'Riordan and Steven Oud and Ashish Panigrahi and Chae-Yeun Park and Daniel Polatajko and Nicolás Quesada and Chase Roberts and Nahum Sá and Isidor Schoch and Borun Shi and Shuli Shu and Sukin Sim and Arshpreet Singh and Ingrid Strandberg and Jay Soni and Antal Száva and Slimane Thabet and Rodrigo A. Vargas-Hernández and Trevor Vincent and Nicola Vitucci and Maurice Weber and David Wierichs and Roeland Wiersema and Moritz Willmann and Vincent Wong and Shaoming Zhang and Nathan Killoran},
journal= {arXiv preprint arXiv:1811.04968},
year = {2022}
}
备注
Code available at https://github.com/XanaduAI/pennylane/ . Significant contributions to the code (new features, new plugins, etc.) will be recognized by the opportunity to be a co-author on this paper