A layer-wise Gaussian Process model estimates DNN training energy from measured probe networks, reducing MAPE from about 40% to about 10% versus FLOPs-based estimation.
Memory- Efficient DNN Training on Mobile Devices
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
THOR: A Generic Energy Estimation Approach for On-Device Training
A layer-wise Gaussian Process model estimates DNN training energy from measured probe networks, reducing MAPE from about 40% to about 10% versus FLOPs-based estimation.