"Revisiting Token Dropping Strategy in Efficient BERT Pretraining."

Qihuang Zhong et al. (2023)

Details and statistics

DOI: 10.48550/ARXIV.2305.15273

access: open

type: Informal or Other Publication

metadata version: 2024-06-12

a service of  Schloss Dagstuhl - Leibniz Center for Informatics