Paddle Quantum: A deep learning platform that empowers Quantum Computing

0 min read

For deep learning enthusiasts, Paddle Quantum comes bearing good news. Paddle Quantum consists of a set of quantum machine learning toolkits that can help scientists and developers build and train quantum neural network models.

Deep dive into the deep learning platform

Baidu established the Institute for Quantum Computing in 2018 with the aim to bridge the gap between quantum computing and AI. With the application of quantum mechanical features like superposition, entanglement, and interferences in information processing tasks, quantum computing can supercharge AI applications.

Paddle Quantum targets at quantum machine learning, quantum chemical simulation, and quantum combinatorial optimization. PaddlePaddle can reduce the number of layers in the quantum computing network by 50%, making the approach very flexible. Additionally, the max-cut problem, which has applications in theoretical physics, VLSI design, etc., can be solved by Baidu’s novel implementation of Quantum Approximate Optimization Algorithm (QAOA).

The newest version of PaddlePaddle

There has been an addition of 39 new algorithms to the library in the newest version of PaddlePaddle. There are now a total of 146 algorithms and over 200 pre-trained models.

• Paddle.js allows developers to use AI directly in the browser or in apps like Wechat, Baidu Chat, etc.

• PaddleClas development toolkit allows for image classification with a pre-trained model trained on 55 million images with over 100,000 classes.

• Parakeet is a text-to-speech toolkit consisting of WaveFlow which can provide fast, highly efficient and quality audio synthesis.

• PLSC allows for large scale image classification model training tasks with the use of parallel strategy.

• Paddle X provides a full-cycle for processing from data pre-processing to model inferencing.

• PaddleCloud allows to quickly access computing resources on the cloud and perform deep learning tasks with one click.

• EasyData has a drag and drop interface to collect data, data labelling, data cleaning, and enhancement.

Baidu is also working in joint collaboration with leading hardware manufacturers like NVIDIA, Huawei, MediaTek, etc. to build a hardware cooperative ecosystem.