• search
    search
  • reviewers
    reviewers
  • feeds
    feeds
  • assignments
    assignments
  • settings
  • logout

MS-BERT: A Multi-layer Self-distillation Approach for BERT Compression Based on Earth Mover's Distance.

Jiahui HuangBin CaoJiaxing WangJing Fan
Published in: CollaborateCom (2) (2021)
Keyphrases
  • data mining
  • multi layer
  • feed forward neural networks
  • neural network
  • error back propagation
  • neural nets
  • single layer
  • image compression
  • information technology
  • multiple layers
  • group sparse coding