Lite Ye
Software Engineer at Meta
Boston, Massachusetts, United States
Summary
🤩
Rockstar🎓
Top SchoolLite Ye is a software engineer with 12 years of experience specializing in machine learning systems and compiler work, currently building ML compiler infrastructure at Meta. Previously a staff ML systems engineer at OctoAI/NVIDIA and an engineer at Google and TripAdvisor, they bridge research-quality ML algorithms with production-grade tooling. Their open-source contributions include core ML improvements to pylearn2—adding learning rules like AdaGrad and refactoring DBM sampling—and portability fixes to Apache TVM’s test infra, reflecting attention to both algorithmic correctness and cross-platform robustness. Trained with an M.S. in Computer Science from UMass Amherst and a BS in Mechanical Engineering from Shanghai Jiao Tong, they combine rigorous CS foundations with engineering practicality. Known for quietly improving testability and backward-compatibility in widely used projects, they favor changes that make complex ML systems more maintainable in heterogeneous environments.
12 years of coding experience
7 years of employment as a software developer
Master of Science (M.S.) Computer Science, Master of Science (M.S.) Computer Science at University of Massachusetts AmherstBachelor of Science (BS) Mechanical Engineering, Bachelor of Science (BS) Mechanical Engineering at Shanghai Jiao Tong University
Chinese
Github Skills (20)
python10
scripting10
machine-learning10
bash10
cicd10
script10
deep-learning10
sh10
shell10
theano10
sgd9
build-system9
test-automation9
nixos8
testing8
Programming languages (13)
Github contributions (5)
apache/tvm
Feb 2022 - Nov 2022
Open deep learning compiler stack for cpu, gpu and specialized accelerators
Role in this project:
DevOps Engineer
DevOps EngineerContributions:99 reviews, 26 commits, 56 PRs in 9 months
Contributions summary:Lite's commits primarily focused on improving the test infrastructure within the Apache TVM repository. They updated various shell scripts used in the testing process, specifically targeting script execution on systems without /bin/bash, such as NixOS. Their changes involved replacing shebangs with `/usr/bin/env bash` and switching from `bash -e` to `set -e` for error handling, thereby enhancing the portability and robustness of the test suite.
metalvulkancompilertensoropencl
lisa-lab/pylearn2
Oct 2014 - Dec 2014
Warning: This project does not have any current developer. See bellow.
Role in this project:ML Engineer
ML EngineerContributions:11 commits in 2 months
Contributions summary:Lite primarily contributed to the core machine learning aspects of the `pylearn2` repository. Their work involved fixing issues within the testing framework, specifically addressing initialization issues in model subclasses and removing deprecated API usage. They also added new learning rules like AdaGrad and made related code adjustments, demonstrating a focus on improving and expanding the library's machine learning capabilities. The user also removed the 'mcmc_steps' parameter, which likely involved code refactoring related to the sampling procedures in the DBM model.
javascripttypescript
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.