PaddlePaddle is an open source deep learning framework provided by Baidu that enables developers and companies to implement their AI ideas safely and quickly. It is being used by more than 30 of Baidu’s offline and online products and services, covering sectors from search to finance to health.

  • A new generation of deep learning frameworks: PaddlePaddle is a new generation of deep learning framework based on “deep learning programming language”. While maintaining a competitive performance, it greatly enhances the framework’s ability to express models and can describe any potential models that may appear.
  • More friendly to large-scale calculations: PaddlePaddle excels in distributed computing through a variety of large-scale computing operations within Baidu. By using EDL technology, it can save a lot of computing resources, and can also support the training of large-scale sparse models.
  • Provides a visualization tool for Deep Learning: Visual DL helps developers easily observe overall training performance and statistics data, such as accuracy/loss metrics, parameter distribution, image/audio sampling and model graph via ONNX. It is an efficient tool to help developers to more easily debug and develop on a deep learning framework.

At Baidu, PaddlePaddle has been deployed into products and services with a vast number of users, including ad click-through rate (CTR) prediction, large-scale image classification, optical character recognition(OCR), search ranking, computer virus detection, recommendation, etc. It is widely utilized in products at Baidu.

Company Type: Enterprise R&D

Region: Asia

Industry Category: Technology

Toolkit: PArallel Distributed Deep LEarning