DTO-EE jointly tunes subtask offloading probabilities and early exit confidence thresholds, reducing simulated response delay by 21-41% while preserving accuracy.
State-of-the-art in smart streetlight systems: a review,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.DC 1years
2024 1verdicts
CONDITIONAL 1roles
background 1polarities
background 1representative citing papers
citing papers explorer
-
Collaborative Inference for Large Models with Task Offloading and Early Exiting
DTO-EE jointly tunes subtask offloading probabilities and early exit confidence thresholds, reducing simulated response delay by 21-41% while preserving accuracy.