MotivationManual discrete prompts have several drawbacks, which contain unstable performance(eg: changing a single word causes substantial drow) . To adress the problem, we propose a novel method whic

Large Language Model Lilian Weng

MotivationFine-tuning has to train and store a full copy of the model parameter, which leads to high compution complexity and large storing resource. To address the problem, Prefix-tuning was proposed