Frozenmad
Software Engineer at THU
Beijing, China
Summary
🤩
RockstarFrozenmad is a software engineer based in Beijing with eight years of hands-on experience building backend systems and machine learning tooling. They contribute to notable open-source projects—helping add interpretability tutorials for BERT models in Microsoft's nlp-recipes and improving AutoGL's packaging, docs, and dataset pipelines—demonstrating an emphasis on reproducible ML and developer experience. Comfortable bridging DevOps and backend work, they’ve streamlined documentation workflows (Read the Docs) and PyPI releases, signaling practical deployment experience beyond core model code. Their contributions blend practical code, documentation, and configuration changes, showing a focus on maintainability and onboarding. Colleagues can expect an engineer who pairs model-level insight with the infrastructure know-how to take ML work from prototype to reproducible package.
8 years of coding experience
Github Skills (15)
configuration-management10
machine-learning10
pytorch10
nlp10
deep-learning10
pypi10
python10
natural-language-processing10
bert10
documentation10
automl9
deeplearning-ai8
plaintext8
graph-neural-network8
pre-trained-model4
Programming languages (6)
Github contributions (5)
THUMNLab/AutoGL
Nov 2020 - Apr 2022
An autoML framework & toolkit for machine learning on graphs.
Role in this project:
Backend Developer & DevOps Engineer
Backend Developer & DevOps EngineerContributions:2 releases, 25 reviews, 306 commits in 1 year 4 months
Contributions summary:Frozenmad's commits primarily focused on documentation updates and configuration file modifications, suggesting a role in maintaining and improving the project's infrastructure and documentation. Specifically, the user updated the configuration file for Read the Docs, which streamlined the documentation process. Additionally, the user contributed to the setup and maintenance of the PyPI package, indicating familiarity with the deployment and distribution aspects of the project. Furthermore, the user refactored code related to datasets, showcasing their understanding of data processing within the framework.
pytorchdata-sciencedeep-learningneural-architecture-searchhyper-parameter-optimization
microsoft/nlp-recipes
Jun 2019 - Jul 2019
Natural Language Processing Best Practices & Examples
Role in this project:ML Engineer
ML EngineerContributions:6 commits in 1 month
Contributions summary:Frozenmad contributed to the `nlp-recipes` repository by adding an explanation component for NLP models. They created a tutorial using a pre-trained BERT model, demonstrating how to interpret and explain a saved PyTorch model. The user implemented code to prepare input data, define the model, and optimize an interpreter to obtain sigma values, which visualize the results. Furthermore, the user has added documentation and removed files.
natural-language-understandingnluword-embeddingslanguage-processingdialogue-systems
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.