Multitask Learning: A general model for NLP?

Slides

• Sequence tagging
  NER; aspect specific sentiment

• Text classification
  dialogue state tracking; sentiment classification

• Seq2Seq
  machine translation; summerization; question answering

• The Natural Language Deathlon(decaNLP)

  • choose from Question/Context/External vocabulary
  • training strategies
    • Fully Joint
    • Anti-Curricaculum Pre-training（train big dataset first than smaller dataset）