Training and inference of large language models using 8-bit floating point.
Sergio P. PerezYan ZhangJames BriggsCharlie BlakeJosh Levy-KramerPaul BalancaCarlo LuschiStephen BarlowAndrew William FitzgibbonPublished in: CoRR (2023)
Keyphrases
- language model
- floating point
- language modeling
- n gram
- fixed point
- document retrieval
- language modelling
- query expansion
- retrieval model
- probabilistic model
- speech recognition
- information retrieval
- ad hoc information retrieval
- test collection
- statistical language models
- sparse matrices
- context sensitive
- vector space model
- training set
- instruction set
- bayesian networks
- relevance model
- translation model
- floating point arithmetic
- language models for information retrieval
- smoothing methods
- language model for information retrieval
- language modeling framework
- query specific
- pseudo relevance feedback