SPDF: Sparse Pre-training and Dense Fine-tuning for Large Language Models.
Vithursan ThangarasaAbhay GuptaWilliam MarshallTianda LiKevin LeongDennis DeCosteSean LieShreyas SaxenaPublished in: CoRR (2023)
Keyphrases
- language model
- fine tuning
- language modeling
- fine tuned
- document retrieval
- viable alternative
- n gram
- probabilistic model
- information retrieval
- speech recognition
- language modelling
- query expansion
- retrieval model
- statistical language models
- test collection
- smoothing methods
- ad hoc information retrieval
- language models for information retrieval
- context sensitive
- document ranking
- pseudo relevance feedback
- relevance model
- query terms
- translation model
- term dependencies
- high dimensional
- vector space model
- retrieval effectiveness
- text retrieval
- generative model
- co occurrence
- document length
- training set
- search engine