Efficient Model Design

Fri, 06 Nov 2026 00:00:00 +0000

My work in neural architecture search focuses on training-free search, hardware-aware model design, and efficient architectures for vision and language models.

Tiny-NAS is a distributed, full-stack framework for zero-shot neural architecture search, with hardware-latency prediction for GPU and mobile deployment. Zen-NAS is a fast, training-free method for finding models that balance predictive accuracy and inference efficiency.

Open Source | Ming Lin

Efficient Model Design