Hung Le
Member Of Technical Staff at Stealth
United States
Summary
🤩
Rockstar🎓
Top SchoolHung Le is a research-driven machine learning engineer with 9 years of experience spanning industry labs and applied research roles at Salesforce, Meta, Amazon, and A*STAR. He holds a PhD in Computer Science from Singapore Management University and has focused on multimodal language-vision models, contributing to notable open-source projects like LAVIS by integrating audio-visual dialogue and video inputs. At Salesforce and Amazon he translated research prototypes into production-ready components, and his GitHub work reflects a practical knack for dataset engineering, processor adaptation, and evaluation metrics. Now a Member of Technical Staff at a stealth startup, he blends deep academic rigor with hands-on implementation to push language-vision intelligence toward real-world applications.
9 years of coding experience
6 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Singapore Management UniversityHigh School Diploma, High School Diploma at Nguyen Thuong Hien High School For The Gifted, HCMC, Vietnam
Vietnamese, English
Github Skills (8)
computer-vision10
pytorch10
machine-learning10
deep-learning10
python10
natural-language-processing10
image-annotation9
image-text9
Programming languages (3)
Github contributions (5)
salesforce/LAVIS
Aug 2022 - Oct 2022
LAVIS - A One-stop Library for Language-Vision Intelligence
Role in this project:
ML Engineer
ML EngineerContributions:12 commits, 2 PRs, 2 pushes in 2 months
Contributions summary:Hung focused on integrating and updating the AVSD (Audio Visual Scene-aware Dialogue) task within the LAVIS framework, specifically modifying and adding code related to dialogue processing, dataset integration, and evaluation metrics. The contributions involved creating and adapting datasets for dialogue tasks and modifying existing processors. The user also updated the model to accommodate video input and related parameters.
vision-frameworksalesforcemultimodal-datasetsvision-and-languagelanguage-model
salesforce/CodeRL
Jul 2022 - Nov 2022
This is the official code for the paper CodeRL: Mastering Code Generation through Pretrained Models and Deep Reinforcement Learning (NeurIPS22).
Contributions:36 commits, 25 pushes, 19 comments in 4 months
pytorchcodegenerationcode-generationlanguagemodelreinforcementlearning
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.