Rethinking Channel Dimensions to Isolate Outliers for Low-bit Weight Quantization of Large Language Models.
Jung Hwan HeoJeonghoon KimBeomseok KwonByeongwook KimSe Jung KwonDongsoo LeePublished in: CoRR (2023)
Keyphrases
- language model
- language modeling
- n gram
- probabilistic model
- speech recognition
- information retrieval
- language modelling
- uniform quantization
- document retrieval
- test collection
- query expansion
- retrieval model
- statistical language models
- context sensitive
- ad hoc information retrieval
- language models for information retrieval
- smoothing methods
- query terms
- bit budget
- translation model
- relevance model
- bayesian networks
- document ranking
- query specific
- vector space model
- okapi bm
- text mining
- data points