ResQ: Mixed-Precision Quantization of Large Language Models with Low-Rank Residuals.

Utkarsh Saxena, Sayeh Sharify, Kaushik Roy, Xin Wang

Browse the full ICML paper archive.