Data Level Lottery Ticket Hypothesis for Vision Transformers

doi:doi:10.57702/jdt06eal

Data Level Lottery Ticket Hypothesis for Vision Transformers

The conventional lottery ticket hypothesis (LTH) claims that there exists a sparse subnetwork within a dense neural network and a proper random initial-ization method called the winning ticket, such that it can be trained from scratch to almost as good as the dense counterpart.

BibTex:

@dataset{Xuan_Shen_and_Zhenglun_Kong_and_Minghai_Qin_and_Peiyan_Dong_and_Geng_Yuan_and_Xin_Meng_and_Hao_Tang_and_Xiaolong_Ma_and_Yanzhi_Wang_2024,
    abstract = {The conventional lottery ticket hypothesis (LTH) claims that there exists a sparse subnetwork within a dense neural network and a proper random initial-ization method called the winning ticket, such that it can be trained from scratch to almost as good as the dense counterpart.},
    author = {Xuan Shen and Zhenglun Kong and Minghai Qin and Peiyan Dong and Geng Yuan and Xin Meng and Hao Tang and Xiaolong Ma and Yanzhi Wang},
    doi = {10.57702/jdt06eal},
    institution = {No Organization},
    keyword = {'Lottery Ticket Hypothesis', 'Sparse Subnetwork', 'Vision Transformers'},
    month = {dec},
    publisher = {TIB},
    title = {Data Level Lottery Ticket Hypothesis for Vision Transformers},
    url = {https://service.tib.eu/ldmservice/dataset/data-level-lottery-ticket-hypothesis-for-vision-transformers},
    year = {2024}
}