On Reinforcement Learning and Distribution Matching for Fine-Tuning Language Models with no Catastrophic Forgetting.
Tomasz KorbakHady ElsaharGermán KruszewskiMarc DymetmanPublished in: CoRR (2022)
Keyphrases
- language model
- fine tuning
- reinforcement learning
- language modeling
- probabilistic model
- document retrieval
- n gram
- speech recognition
- language modelling
- retrieval model
- viable alternative
- query expansion
- information retrieval
- mixture model
- fine tune
- statistical language models
- language models for information retrieval
- ad hoc information retrieval
- pseudo relevance feedback
- document ranking
- vector space model
- fine tuned
- context sensitive
- test collection
- language model for information retrieval
- translation model
- smoothing methods
- okapi bm
- query terms
- passage retrieval
- relevance model
- cross lingual