SMILE: A Cost-Effective System for Serving Massive Pretrained Language Models in The Cloud.
Jue WangKe ChenLidan ShouDawei JiangGang ChenPublished in: SIGMOD Conference Companion (2023)
Keyphrases
- cost effective
- language model
- data center
- private cloud
- language modeling
- cloud computing
- n gram
- low cost
- document retrieval
- probabilistic model
- retrieval model
- query expansion
- information retrieval
- cost effectiveness
- speech recognition
- language models for information retrieval
- context sensitive
- test collection
- data analysis
- vector space model
- statistical language models
- translation model
- smoothing methods
- language modelling
- ad hoc information retrieval
- real time
- clustering algorithm
- relevance model
- document length
- word error rate
- pseudo relevance feedback
- retrieval effectiveness
- statistical language modeling
- query terms