LiyuanLucasLiu / LD-Net

Efficient Contextualized Representation: Language Model Pruning for Sequence Labeling
https://arxiv.org/abs/1804.07827
Apache License 2.0
146 stars 13 forks source link

Do you think we can edit transformer to dense structure like LD-Net? #9

Closed guotong1988 closed 5 years ago

guotong1988 commented 5 years ago

https://github.com/divelab/dtn Is the link I want? @LiyuanLucasLiu Thank you very much!

LiyuanLucasLiu commented 5 years ago

Emmm, i think it would be better to discuss such topics via email or IMs, since issues here are supposed to be related to the LD-Net model only. Thanks for your understanding :- )