AAAI2024AlgoOWQ: Outlier-Aware Weight Quantization for Efficient Fine-Tuning and Inference of Large Language Models