Login / Signup
Towards Non-task-specific Distillation of BERT via Sentence Representation Approximation.
Bowen Wu
Huan Zhang
Mengyuan Li
Zongsheng Wang
Qihang Feng
Junhong Huang
Baoxun Wang
Published in:
CoRR (2020)
Keyphrases
</>
natural language sentences
natural language
neural network
genetic algorithm
multiscale
closed form
approximation algorithms
feature representation
approximation error
discourse structure