Tao Lei - Research Scientist at Google

Tao Lei

Research Scientist at Google

Mountain View, California, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Tao Lei is a research scientist at Google Research with nine years of experience building efficient NLP algorithms and leading applied machine learning teams. Previously he scaled NLP research into production as VP of NLP at ASAPP, managing researchers and engineers to advance enterprise-facing language models. He holds an MS and PhD in Computer Science from MIT and a BS from Peking University, blending deep academic rigor with product-oriented impact. Tao contributes to open-source deep learning tooling—most notably optimizing SRU recurrent kernels with CUDA and TorchScript support—to make RNNs competitive with CNNs for high-performance inference. Based in Mountain View, he combines low-level systems chops with large-scale model development to drive both research breakthroughs and deployable solutions.

9 years of coding experience

4 years of employment as a software developer

Master's degree, Computer Science, Master's degree, Computer Science at Massachusetts Institute of Technology

Bachelor of Science - BS, Computer Science, Bachelor of Science - BS, Computer Science at Peking University

Stackoverflow

Stats

1reputation

0reached

0answers

0questions

Github Skills (8)

cuda10

gpu-programming10

pytorch10

recurrent-neural-networks10

deep-learning10

python9

cprogramming-language9

c-language9

Programming languages (1)

Python

Github contributions (5)

asappresearch/sru

Aug 2017 - Jun 2021

Training RNNs as Fast as CNNs (https://arxiv.org/abs/1709.02755)

Role in this project:

Backend Developer

Contributions:28 releases, 20 reviews, 173 commits in 3 years 10 months

Contributions summary:Tao's commits primarily involve the implementation of CUDA kernels for the SRU (Simple Recurrent Unit) model. These changes focus on optimizing and extending the SRU implementation to support CPU and GPU based inference. The user's contributions include improvements to the CUDA kernels, addressing potential issues with mask padding, and ensuring the code is compatible with Python 3. The changes also include additions to support torchscript for the model.

nlppytorchgruarxivabs

asappresearch/flop

Jan 2020 - Jun 2021

Pytorch library for factorized L0-based pruning.

Contributions:29 commits, 11 PRs, 3 pushes in 1 year 4 months

pytorchpruningpytorch-library

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.

Request Free Trial