TY - RPRT TI - DL-QAT: Weight-Decomposed Low-Rank Quantization-Aware Training for Large Language Models AU - Wenjin Ke AU - Zhe Li AU - Dong Li AU - Lu Tian AU - Emad Barsoum PY - 2025 UR - https://arxiv.org/abs/2504.09223 ID - 2504.09223 ER -