Tung Le
Senior Research Scientist at IBM Research - Tokyo
Chiba Prefecture, Japan
Summary
👤
Senior🎓
Top SchoolTung Le is a Senior Research Scientist at IBM Research - Tokyo with 11 years of experience at the intersection of high-performance computing, programming languages, and deep learning. His work spans from research and postdoctoral projects to staff and senior roles, driving compiler-level innovations and practical ML tooling. He has contributed to the onnx-mlir project, implementing new operator lowerings and improving dynamic-dimension handling to bridge ONNX models with MLIR infrastructure. With a Ph.D. from the National Institute of Informatics and engineering roots in Hanoi, he blends rigorous academic foundations with hands-on systems development. Based in Chiba, Japan, Tung is especially adept at turning compiler theory into production-ready optimizations for AI workloads. Colleagues rely on him for deep technical insight into operator semantics and performant backend implementations.
11 years of coding experience
14 years of employment as a software developer
Master of Science (M.Sc.), Information Technology, Master of Science (M.Sc.), Information Technology at Hanoi University of Science and TechnologyPh.D, Computer Science, Ph.D, Computer Science at National Institute of Informatics, Tokyo, Japan
Github Skills (12)
tensorrt10
tensor10
tensorflow10
onnx10
operation10
mlr10
cprogramming-language9
c-language9
deep-learning8
cluster-computing8
scientific-computing8
parallel-computing8
Programming languages (3)
Github contributions (5)
onnx/onnx-mlir
Mar 2020 - Jan 2023
Representation and Reference Lowering of ONNX Models in MLIR Compiler Infrastructure
Role in this project:
Back-end Developer & AI/ML Engineer
Back-end Developer & AI/ML EngineerContributions:1451 reviews, 259 commits, 755 PRs in 2 years 10 months
Contributions summary:Tung primarily contributed to the implementation of operators for the ONNX framework, including the creation of new lowering rules. They focused on enhancing the functionality of the framework by incorporating new operators like `NonZero`, expanding the capabilities of existing ones such as `Concat`, and refining the handling of dynamic dimensions. The user also developed and optimized the code, integrating the new operations with the existing infrastructure and correcting potential errors.
pytorchrepresentationdeep-learningmlironnx-models
tungld/ONNF
Jan 2020 - Feb 2020
Open Neural Network Frontend
Contributions:195 pushes, 34 branches in 1 month
neural-networkfrontend
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.