BriefGPT.xyz
Oct, 2023
Point-PEFT: 3D预训练模型的参数高效微调
Point-PEFT: Parameter-Efficient Fine-Tuning for 3D Pre-trained Models
HTML
PDF
Ivan Tang, Eric Zhang, Ray Gu
TL;DR
通过Point-PEFT框架,我们能够在只使用可训练参数的5%的情况下,实现优于完全微调的性能,充分展示了我们方法的高效性和有效性。
Abstract
The popularity of pre-trained large models has revolutionized
downstream tasks
across diverse fields, such as language, vision, and multi-modality. To minimize the adaption cost for
downstream tasks
, many
→