Jiecao Yu
Senior Staff Research Scientist at Facebook
Ann Arbor, Michigan, United States
Summary
🤩
Rockstar🎓
Top SchoolJiecao Yu is a Senior Staff Research Scientist at Meta with a decade of experience building high-performance AI infrastructure and co-designing software/hardware for large language model deployments. Based in Ann Arbor, he progressed through research roles at Facebook (now Meta) from intern to senior staff, blending systems engineering with ML research to optimize model inference and training stacks. His open-source contributions include work on FBGEMM—improving sparse matrix kernels and SparseAdagrad performance—and a PyTorch XNOR-Net implementation, reflecting deep expertise in numerical kernels and efficiency-focused model engineering. He holds advanced degrees in Computer Engineering from the University of Michigan and brings an uncommon combination of low-level optimization skills and production ML systems experience to scale LLM infra.
10 years of coding experience
6 years of employment as a software developer
Computer Engineering, Computer Engineering at University of Michigan - Rackham Graduate SchoolChunhui Middle SchoolAdvanced Honor Class of Engineering Education (ACEE) ACEE Minor Chu Kochen Honors College (CKC), Advanced Honor Class of Engineering Education (ACEE) ACEE Minor Chu Kochen Honors College (CKC) at Zhejiang University
Github Skills (25)
algorithm10
algorithms10
pytorch10
c-language10
fba10
numerical10
emm10
sparse-matrix10
machine-learning10
mask-rcnn10
gemfire10
deep-learning10
f10
numerical-methods10
performance-optimization10
Programming languages (4)
Github contributions (5)
jiecaoyu/XNOR-Net-PyTorch
Sep 2017 - Mar 2019
PyTorch Implementation of XNOR-Net
Role in this project:
ML Engineer
ML EngineerContributions:66 commits, 54 pushes, 1 branch in 1 year 6 months
Contributions summary:Jiecao primarily focused on implementing and modifying a PyTorch implementation of XNOR-Net, a binarized neural network. Their contributions include adding and modifying code related to the CIFAR-10 dataset, including data loading and model definition. They also fixed gradient-related issues and updated the model to be compatible with the new pytorch version.
pytorchefficientnetvisual-recognitionpytorch-implementationresnet
pytorch/FBGEMM
Jan 2020 - Jul 2022
FB (Facebook) + GEMM (General Matrix-Matrix Multiplication) - https://code.fb.com/ml-applications/fbgemm/
Role in this project:ML Engineer
ML EngineerContributions:6 commits, 9 PRs, 2 comments in 2 years 6 months
Contributions summary:Jiecao contributed to the fbgemm repository by implementing and optimizing sparse matrix operations, specifically related to the SparseAdagrad algorithm. Their work involved adding interfaces, refactoring existing code, and improving the performance of kernel functions. They also participated in benchmark testing to assess and compare the speed of different implementations. These changes reflect a focus on enhancing the performance and functionality of core numerical algorithms within the library.
matrix-multiplicationfacebookmultiplicationmatrixml-applications
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.