Gasoonjia
From Beijing, live in Santa Clara, United Kingdom
Summary
🤩
RockstarShengjia Zhao is Chief Scientist at Meta Superintelligence Labs with nine years of experience bridging cutting‑edge research and production ML systems. Previously a Member of Technical Staff at OpenAI and a PhD candidate at Stanford, he combines deep academic training with hands‑on engineering to ship large‑scale model work. He has contributed to PyTorch core—improving tensor semantics and the numerical debugger—highlighting a practical focus on reliability and debuggability in ML frameworks. Based in San Francisco, Shengjia blends research rigor with systems pragmatism to move ideas from papers into reliable tooling and infrastructure. Colleagues know him for digging into subtle numerical and control‑flow issues that make large models tractable in real deployments.
9 years of coding experience
Github Skills (15)
tensorrt10
debugging10
pytorch10
machine-learning10
debug10
tensor10
tensorflow10
python10
operation10
numpy9
autograd9
neural-network9
deeplearning-ai9
deep-learning9
gpu8
Programming languages (7)
Github contributions (5)
pytorch/pytorch
Mar 2018 - Feb 2025
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
ML Engineer
ML EngineerContributions:4 reviews, 15 PRs, 56 pushes in 6 years 11 months
Contributions summary:Gasoonjia primarily contributed to the PyTorch library, focusing on improvements related to tensor operations and the numerical debugger. They addressed issues with `dim_order`, ensuring its correct behavior and raising runtime errors for ambiguous cases. Additionally, the user enhanced the numerical debugger by supporting control flow statements and maintaining debug handler consistency through decomposition steps, indicating a focus on debugging and optimization within the PyTorch framework.
pythongpu-accelerationdeep-learninggpunumpy
Gasoonjia/executorch-1
Oct 2023 - Apr 2025
End-to-end solution for enabling on-device AI across mobile and edge devices for PyTorch models
Contributions:576 pushes, 119 branches in 1 year 5 months
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.