Search

PArallel Distributed Deep LEarning: Machine Learning Framework from Industrial Practice （『飞桨』核心框架，深度学习&机器学习高性能单机、分布式训练和跨平台部署）

PaddlePaddle High Performance Deep Learning Inference Engine for Mobile and Edge (飞桨高性能深度学习端侧推理引擎）

The ripgrep of AI context: a zero-dependency C++23 CLI + MCP server for coding agents. Find what you want without reading the repo, then check you built what you meant — blast radius, tests-to-run, quality deltas. Signatures at 74.7% fewer bytes than bodies; every guess labelled, every loss published. Paddle out with a map.

纯 C# PP-OCRv6 推理库：多平台手写 Kernel / GEMM 等算子、超高性能、低内存需求、超高准确率。 自带托管 ONNX 解释器，不依赖 Paddle Inference、ONNX Runtime 或 OpenCV 原生库。