Fangzhou Mu
California, United States
Summary
👤
SeniorFangzhou Mu is a research scientist specializing in generative and multimodal AI with a decade of experience building data infrastructure, curation methodologies, and evaluation pipelines. Based in California, he has driven production-quality MLLM serving and Day-0 model support at NVIDIA and now advances generative AI research at Meta. His work spans optimizing inference on cutting-edge GPUs, shipping containers and endpoints for frontier models, and inventing novel image restoration and single-image 3D stylization techniques presented at CVPR. Trained with an MS and PhD from the University of Wisconsin–Madison and a top-ranked BS from Zhejiang University, he blends deep academic rigor with fast-paced production delivery. A practical innovator, Fangzhou pairs low-level performance engineering with creative model design to bridge research and large-scale deployment.
10 years of coding experience
Github Skills (38)
alpha10
meta-learning10
maml10
action-localization10
super-resolution10
self-attention10
few-shot-learning10
generative-adversarial-network9
face9
representation-learning9
classification9
machine-learning9
language-modeling9
3d9
sklearn9
Programming languages (4)
Github contributions (5)
fmu2/Wasserstein-BiGAN
Dec 2019 - Dec 2019
Wasserstein BiGAN (Bidirectional GAN trained using Wasserstein distance)
Contributions:19 commits, 21 pushes, 1 branch in 1 day
deep-learninggenerative-adversarial-networkganwasserstein-gan
shirleyzhu233/realNVP
Jan 2019 - Feb 2019
PyTorch implementation of realNVP
Contributions:30 commits, 2 PRs, 22 pushes in 28 days
pytorch
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.