ToEx: Accelerating Generation Stage of Transformer-Based Language Models via Token-Adaptive Early Exit.
Myeonggu KangJunyoung ParkHyein ShinJaekang ShinLee-Sup KimPublished in: IEEE Trans. Computers (2024)
Keyphrases
- language model
- language modeling
- n gram
- document retrieval
- information retrieval
- probabilistic model
- language modelling
- retrieval model
- query expansion
- speech recognition
- query terms
- language models for information retrieval
- statistical language models
- language model for information retrieval
- pseudo relevance feedback
- context sensitive
- ad hoc information retrieval
- test collection
- translation model
- document ranking
- query specific
- smoothing methods
- relevance model
- vector space model
- document length
- word error rate
- okapi bm
- co occurrence