Sheng Xu
Research Scientist at Soochow University
Suzhou City, Jiangsu, China
Summary
🤩
RockstarSheng Xu is a research scientist and NLP practitioner based in Suzhou with 11 years of experience building and optimizing transformer-based models. He contributes practical tutorials and hands-on implementations—such as a popular quick-start Transformers guide—and has applied PyTorch to tasks ranging from FashionMNIST classification to AFQMC pairwise similarity, NER, translation, and summarization. Sheng blends research rigor with engineering focus, owning model training, evaluation, and optimization across diverse NLP datasets. Notably, his open-source work emphasizes making advanced transformer workflows accessible to practitioners, reflecting a commitment to reproducible, applied research.
11 years of coding experience
Github Skills (11)
transformers10
pytorch10
machine-learning10
nlp10
bert10
classification10
natural-language-processing9
translation7
summarization7
named-entity-recognition7
tensorflow4
Programming languages (4)
Github contributions (5)
jsksxs360/How-to-use-Transformers
Sep 2022 - Nov 2022
Transformers 库快速入门教程
Role in this project:
ML Engineer
ML EngineerContributions:44 commits, 7 PRs, 73 pushes in 2 months
Contributions summary:Sheng demonstrates expertise in machine learning model development, specifically within the domain of natural language processing and transformers. They implemented and optimized a neural network for FashionMNIST classification using PyTorch. The user further added and trained various models, including those for pairwise classification similarity (AFQMC dataset), named entity recognition (PeopleDaily dataset), translation, and summarization. The user's contributions showcase a focus on model training and evaluation across several NLP tasks.
nlppytorchtransformerslanguage-modelingthai
ZHResearch/ZHEvent.github.io
Sep 2020 - Oct 2022
事件小分队论文笔记
Contributions:78 commits, 77 pushes, 1 branch in 2 years 1 month
java
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.