CSS-LM: A Contrastive Framework for Semi-Supervised Fine-Tuning of Pre-Trained Language Models.
YuSheng SuXu HanYankai LinZhengyan ZhangZhiyuan LiuPeng LiJie ZhouMaosong SunPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2021)
Keyphrases
- language model
- probabilistic model
- fine tuning
- language modeling
- semi supervised
- n gram
- speech recognition
- information retrieval
- test collection
- document retrieval
- relevance model
- language modelling
- pre trained
- labeled data
- semi supervised learning
- retrieval model
- query expansion
- pairwise
- knn
- statistical language models
- language modeling framework
- term dependencies
- document ranking
- machine learning
- learning algorithm
- bayesian networks