Login / Signup

How to Distill your BERT: An Empirical Study on the Impact of Weight Initialisation and Distillation Objectives.

Xinpeng WangLeonie WeissweilerHinrich SchützeBarbara Plank
Published in: CoRR (2023)
Keyphrases
  • three dimensional
  • databases
  • fully automatic
  • search engine
  • data structure
  • lower bound
  • expert systems
  • multiple objectives
  • weighting scheme
  • weight assignment