Offline Inference Engine

in Youdao, 2021

This project shows Youdao’s Offline Inference Engine, which deploys deep models on various devices using self-developed controller based on open source frameworks. It has successfully run on all Youdao’s smart hardware, like the Youdao DictPen Series, Youdao SuperDict and Youdao Translator.

Here are the detailed reports:

  1. 网易有道开源EMLL:高性能端侧机器学习计算库,大幅提高计算性能
Bin Xu
Bin Xu
Algorithm Engineer

My research interests include AI for science, spatio-temporal data analysis, large language model and computer vision.