Accelerating BERT inference with GPU-efficient exit prediction.

Lei Li Chengyu Wang Minghui Qiu Cen Chen Ming Gao Aoying Zhou

Published in: Frontiers Comput. Sci. (2024)

Keyphrases

prediction accuracy
real time
cost effective
computationally expensive
inference process
image processing
computationally efficient
parallel architectures
feature selection
data structure
video sequences
lightweight
probabilistic inference
efficient learning
prediction algorithm
graphics hardware