Tao Lv
AI Frameworks Engineer at 英特尔
Songjiang District, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolTao Lv is an AI Frameworks Engineer with nine years of experience designing and optimizing deep learning runtimes, currently focused on oneDNN Graph development at Intel in Shanghai. He brings practical low-level expertise in CPU/GPU tensor libraries and matrix kernels, evidenced by contributions to mshadow and MXNet where he optimized batch GEMM paths and hardened global pooling tests. Tao has improved cross-platform compatibility and testing in oneDNN, adding backend integrations and fixing Windows build issues—work that underpins reliable production-grade inference and training stacks. With a Master’s in Signal and Information Processing and a background in performance engineering on Intel architectures, he combines algorithmic insight with pragmatic engineering to squeeze real-world performance from ML frameworks.
9 years of coding experience
9 years of employment as a software developer
Bachelor of Science (B.S.), Computer Science and Technology, Bachelor of Science (B.S.), Computer Science and Technology at Xidian University
Chinese, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (30)
c-language10
matrix10
python10
testing10
graph10
mxnet10
cluster-computing10
standard-library10
mat10
oneapi10
deep-learning10
parallel-computing10
deep-neural-networks10
cuda10
c-library10
Programming languages (6)
Github contributions (5)
uxlfoundation/oneDNN
Mar 2018 - Jan 2023
oneAPI Deep Neural Network Library (oneDNN)
Role in this project:
Back-end Developer
Back-end DeveloperContributions:191 reviews, 713 commits, 101 PRs in 4 years 10 months
Contributions summary:Tao contributed to the oneDNN library by implementing and modifying core API components. Their work involved adding new features like sentinel values to enums, fixing build errors on Windows, and integrating a DNNL backend for graph functionalities. The user's contributions also included the addition of graph tests and the creation of a fake backend. These changes enhance the library's functionality, compatibility, and testing capabilities.
bfloat16sse41avx512openmpopencl
apache/mxnet
Feb 2018 - Sep 2020
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:ML Engineer & Test Automation Engineer
ML Engineer & Test Automation EngineerContributions:4 releases, 5 reviews, 76 commits in 2 years 7 months
Contributions summary:Tao's commits primarily revolve around testing and improving the `mxnet` deep learning framework's core functionalities. They focused on validating the global pooling operations across different implementations (CPU and GPU), as well as in the test suite. The commits include adding new test cases, checking for padding sizes in various implementations, and addressing compilation issues within the CI environment. These improvements highlight their role in ensuring the reliability and correctness of the framework's performance.
pythonschedulerdataflowmutationdata-science
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.