Cost-Efficient Large Language Model Serving for Multi-turn Conversations with CachedAttention.
Bin GaoZhuomin HePuru SharmaQingxuan KangDjordje JevdjicJunbo DengXingkun YangZhou YuPengfei ZuoPublished in: USENIX ATC (2024)
Keyphrases
- language model
- cost efficient
- language modeling
- n gram
- probabilistic model
- document retrieval
- retrieval model
- speech recognition
- information retrieval
- test collection
- language modelling
- query terms
- query expansion
- mixture model
- statistical language models
- context sensitive
- ad hoc information retrieval
- smoothing methods
- pseudo relevance feedback
- translation model
- vector space model
- document ranking
- relevance model
- retrieval effectiveness
- language model for information retrieval
- word error rate
- pseudo feedback
- relevant documents