Paddle.js is a web project for Baidu PaddlePaddle, which is an open source deep learning framework running in the browser. Paddle.js can either load a pre-trained model, or transforming a model from paddle-hub with model transforming tools provided by Paddle.js. It could run in every browser with WebGL/WebGPU supported.
- paddlejs-core, the core part of the Paddle.js ecosystem, which is responsible for the operation of the inference process of the entire engine.
- paddlejs-converter, model transformation tool for Paddle.js, convert PaddlePaddle models (also known as fluid models) into a browser-friendly format.
- paddlejs-models, model projects, supply flexible low-level APIs for users to implement their AI scenario.
- paddlejs-examples, Paddle.js AI examples
- paddlejs-mediapipe, tools for live and streaming media, support webrtc camera and a lightweight opencv package
- paddlejs-backend-webgl, webgl backend, the main backend for Paddle.js, ops supported
- paddlejs-backend-webgpu，webgpu backend, an experimental backend, WebGPU draft, ops supported
- paddlejs-backend-cpu, cpu backend, ops supported
- paddlejs-backend-nodegl, nodegl backend, a backend in Node.js for Paddle.js, using the webgl backend ops. ops supported
- PC: Chrome, Safari, Firefox
- Phone: Baidu App , Chrome , UC and QQ Browser
- Smartprogram: Baidu Smartprogram, WX Smartprogram
- Platform: macOS, Windows
The conversion script above produces 2 types of files:
- model.json (the dataflow graph and weight manifest file)
- chunk_x.dat (collection of binary weight files)
- Online video tutorial start video
- Questions, reports, and suggestions are welcome through Github Issues!
- Forum: Opinions and questions are welcome at our PaddlePaddle Forum！
- QQ group chat: 696965088