Zhi Zhang
PHD Student at The Hong Kong Polytechnic University
Kowloon, Hong Kong, China
Summary
🤩
RockstarZhi Zhang is a PhD student and software engineer based in Kowloon, Hong Kong, with a decade of hands-on experience in machine learning engineering and distributed training systems. He focuses on practical performance improvements, evidenced by contributions to PyTorch distributed training—fixing training loop bugs and adding support for Apex, Horovod, native torch.distributed, and Slurm-based workflows. His work blends rigorous research-level thinking with production sensibilities, enabling scalable training across clusters. Fluent in both academic and open-source cultures, he pairs a scholarly perspective with a pragmatic motto reflected on his GitHub: continuous self-awareness and capability growth. Colleagues can expect a meticulous engineer who thrives on optimizing complex ML pipelines for real-world throughput and efficiency.
10 years of coding experience
Github Skills (9)
pytorch10
machine-learning10
distributed-training10
python10
horovod9
apex9
slurm8
gpu8
imagenet7
Programming languages (11)
Github contributions (5)
tczhangzhi/pytorch-distributed
Dec 2019 - Aug 2021
A quickstart and benchmark for pytorch distributed training.
Role in this project:
ML Engineer
ML EngineerContributions:10 commits, 21 pushes, 1 branch in 1 year 8 months
Contributions summary:Zhi primarily focused on enhancing distributed training capabilities within the PyTorch framework. Their contributions include fixing loop bugs within training scripts, and adding and refining scripts for different distributed training methods such as Apex, Horovod, and the standard PyTorch `distributed` package. Furthermore, they added a script for distributed training within a Slurm environment. These changes suggest a focus on improving the performance and efficiency of distributed training setups.
distributed-trainingpytorch
tczhangzhi/PyTorch-GANSER
Jul 2021 - May 2022
Official implementation of the paper "GANSER: A Self-supervised Data Augmentation Framework for EEG-based Emotion Recognition"
Contributions:3 commits, 2 pushes, 1 branch in 10 months
data-augmentationemotion-recognition
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.