models
Officially maintained, supported by PaddlePaddle, including CV, NLP, Speech, Rec, TS, big models and so on.
This repository is the official model library for PaddlePaddle, an AI development platform from Baidu. PaddlePaddle is China's most widely used open-source deep-learning framework, comparable in purpose to PyTorch or TensorFlow. The models library is a curated collection of pre-built AI models that developers can download and use in their own applications without training from scratch.
The collection covers a wide range of AI tasks: understanding text, classifying images, detecting objects in photos, segmenting images into regions, recognizing text in images, and synthesizing speech. As of the most recent update noted in the README, the official model count exceeded 600, with an additional 260-plus models contributed by the broader ecosystem of researchers and partners.
The repository organizes its contents into three categories. Official models are production-ready, maintained by the PaddlePaddle team, and designed for use in commercial and industrial applications across fields like finance, manufacturing, and agriculture. Research models are experimental and oriented toward academic work, contributed by research partners. Community models are contributed by external developers and cover a broader range of use cases.
The README is written entirely in Chinese and is aimed at Chinese-speaking developers working in industries that use AI for business automation. No significant setup instructions appear in the README itself, suggesting that users are expected to consult the linked documentation pages for each model category. The library is released under the Apache 2.0 license.