SADA accelerates ODE-based generative model sampling by adaptively combining step skipping and token pruning through a stability criterion, giving about 1.8 times speedup with minor fidelity loss.
U-net: Con- volutional networks for biomedical image segmenta- tion
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
-
SADA: Stability-guided Adaptive Diffusion Acceleration
SADA accelerates ODE-based generative model sampling by adaptively combining step skipping and token pruning through a stability criterion, giving about 1.8 times speedup with minor fidelity loss.