Introducing PQNT — A New Power-Law Quantization Method

_By: Michael Anggi Gilang Angkasa

🌟Why I Built PQNT
As deep learning models grow larger, inference efficiency becomes a serious bottleneck — especially on edge devices, mobile CPUs, and embedded systems. Quantization has become a standard way to accelerate inference and reduce model size (FP32 → INT8 = 4× compression).
For full repository :
https://github.com/Michael-Obs66/pqnt