English

PARQ: Piecewise-Affine Regularized Quantization

Machine Learning 2025-03-21 v1 Optimization and Control

Abstract

We develop a principled method for quantization-aware training (QAT) of large-scale machine learning models. Specifically, we show that convex, piecewise-affine regularization (PAR) can effectively induce the model parameters to cluster towards discrete values. We minimize PAR-regularized loss functions using an aggregate proximal stochastic gradient method (AProx) and prove that it has last-iterate convergence. Our approach provides an interpretation of the straight-through estimator (STE), a widely used heuristic for QAT, as the asymptotic form of PARQ. We conduct experiments to demonstrate that PARQ obtains competitive performance on convolution- and transformer-based vision tasks.

Keywords

Cite

@article{arxiv.2503.15748,
  title  = {PARQ: Piecewise-Affine Regularized Quantization},
  author = {Lisa Jin and Jianhao Ma and Zechun Liu and Andrey Gromov and Aaron Defazio and Lin Xiao},
  journal= {arXiv preprint arXiv:2503.15748},
  year   = {2025}
}
R2 v1 2026-06-28T22:27:38.630Z