Accelerating Pretrained Language Model Inference Using Weighted Ensemble Self-distillation.
Jun KongJin WangXuejie ZhangPublished in: NLPCC (1) (2021)
Keyphrases
- language model
- language modeling
- probabilistic model
- n gram
- query expansion
- document retrieval
- speech recognition
- language modelling
- information retrieval
- retrieval model
- context sensitive
- query terms
- ad hoc information retrieval
- bayesian inference
- vector space model
- smoothing methods
- mixture model
- bayesian networks
- statistical language models
- ensemble methods
- test collection
- web search
- relevance model
- dirichlet prior
- document length
- statistical language modeling
- document ranking
- term weighting
- base classifiers
- retrieval effectiveness
- decision trees