Zeming Lin
Member Of Technical Staff at EvolutionaryScale
New York City Metropolitan Area United States
Summary
🤩
Rockstar🎓
Top SchoolZeming Lin is a Principal Research Scientist based in the NYC area with 12 years of experience building and shipping foundational AI and ML systems for biology and beyond. An early member of the PyTorch team and contributor to high-impact open-source projects like TorchCraft and Facebook Research’s ESM/ESMFold (work published in Science), he blends deep research rigor with pragmatic engineering. He has driven protein language models and structure prediction at scale and cofounded EvolutionaryScale.ai to translate these models into applied biology products. Zeming’s background spans academia and industry—PhD work at NYU, research engineering at Facebook AI Research, and internships at Google—giving him both theoretical depth and production experience. Notably, he contributes to developer tooling and documentation (e.g., improving TorchCraft’s Docker and replay tooling) which reflects a preference for robust, reproducible systems as much as novel models.
12 years of coding experience
7 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at New York UniversityHigh School, High School at Thomas Jefferson High School for Science and TechnologyMaster’s Degree Machine Learning, Master’s Degree Machine Learning at University of Virginia
Chinese, English
Github Skills (19)
markdown10
c-language10
markdown-it10
jekyll10
cprogramming-language10
torch-hub9
lua9
torch-net9
html9
docker8
css8
dockers8
pytorch7
neural-network7
neural7
Programming languages (14)
Github contributions (5)
TorchCraft/TorchCraft
Dec 2016 - Jul 2019
Connecting Torch to StarCraft
Role in this project:
Backend Developer
Backend DeveloperContributions:4 releases, 131 commits, 114 PRs in 2 years 7 months
Contributions summary:Zeming primarily focused on fixing documentation, links, and setup for the TorchCraft project, specifically concerning Docker integration and release version updates. They addressed bugs in the code, improved the robustness of heuristics, and implemented changes related to replayer functionality, including packing neutral IDs for replays and fixing a bug related to handling null frames. Furthermore, they added keyframe functionality to lua.
starcraftminecraftbwapideep-learningreinforcement-learning
Atcold/NYU-DLSP20
Feb 2020 - Sep 2022
NYU Deep Learning Spring 2020
Role in this project:Full-stack Developer
Full-stack DeveloperContributions:2 releases, 47 commits, 44 PRs in 2 years 7 months
Contributions summary:Zeming contributed to the development of the project's website, including implementing features and fixing issues. They worked on theming the website with mdbook and adding functionality like macro support and line numbers to the table of contents. They also made adjustments to the layout files and updated instructions related to the project. Additionally, the user made changes to the formatting of attributions and collaborated on content updates.
pytorchdeep-learningjupyter-notebookneural-netsmachine-learning
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.