Thomas Dao
Machine Learning Engineer
Summary
🤩
RockstarThomas Dao is a Machine Learning Engineer with a decade of experience specializing in deep learning, graph representation learning, reinforcement learning, system design, and blockchain applications. He combines research-minded modeling skills with practical engineering, exemplified by his contributions to an EfficientDet PyTorch implementation where he integrated focal loss and improved dataset/dataloader robustness for object detection. Comfortable moving models toward production, he bridges algorithmic innovation and data engineering to improve training stability and deployment readiness. Known for tackling tricky loss functions and dataset issues, he brings a pragmatic, systems-aware approach to ML projects and open-source collaboration.
10 years of coding experience
Github Skills (4)
computer-vision10
object-detection10
python10
pytorch10
Programming languages (10)
Github contributions (5)
toandaominh1997/EfficientDet.Pytorch
Nov 2019 - Jan 2020
Implementation EfficientDet: Scalable and Efficient Object Detection in PyTorch
Role in this project:
ML Engineer
ML EngineerContributions:90 commits, 38 PRs, 123 pushes in 1 month
Contributions summary:Thomas implemented Focal Loss, a crucial component for object detection models, within the repository's PyTorch framework. They modified the `models/module.py` and `models/losses.py` files to integrate and utilize the focal loss functionality, showcasing a direct involvement in enhancing the model's training process. Further contributions included fixing and modifying the dataset and dataloader to address issues related to dataset handling. These actions directly improved the model's object detection capabilities and its related data processing.
pytorchscalableobject-detectionefficientdetcoco
toandaominh1997/Kaggle
Jun 2018 - Feb 2019
Contributions:55 commits, 39 pushes, 1 branch in 8 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.