1 code implementation • 9 Apr 2024 • Afzal Ahmad, Linfeng Du, Zhiyao Xie, Wei zhang
We present a technique that allows searching for training proxies that reduce the cost of benchmark construction by significant margins, making it possible to construct realistic NAS benchmarks for large-scale datasets.
no code implementations • 5 Mar 2019 • Afzal Ahmad, Muhammad Adeel Pasha
We also design a pipelined and parallel Winograd convolution engine that improves the throughput and power-efficiency while reducing the computational complexity of the overall system.