vla.cpp
The open-source project I'm P.I.C of at VinRobotics - bringing Vision-Language-Action models to local hardware, in the spirit of llama.cpp for robotics.
View on GitHub ↗Havilab
My homelab for edge AI,
build simple and effective code.
I'm Dang-Khanh Nguyen - a software engineer working on edge AI deployment, model optimization, and robotics with VLMs, VLAs, and LLMs. Happy vicuna laboratory is inspired by llama.cpp: small lines of code that bring big things to your own hardware.
About the vicuna behind the lab
I build and ship software where deep learning meets real hardware. My work focuses on designing SDKs for computer-vision models, optimizing neural networks for AI accelerators, and decomposing operations so they run fast and in parallel - getting capable models onto edge devices and robots.
Pinned projects
A few things I'm proud of - small in code, big in scope.
REX
Neural network Representation EXchange - a hands-on study of how different frameworks represent and serialize neural networks.
View on GitHub ↗FaceGen
Multiple Appropriate Facial Reaction Generation - synthesizing diverse listener reactions from a speaker's audio-visual cues. 3rd place at REACT 2024.
View on GitHub ↗Say hello
Always happy to talk edge AI, optimization, and robotics.