PTSBench benchmarks post-training sparsity techniques and model families, finding learning-based allocation and block-wise reconstruction most effective, and attention-based models most sparsity-friendly.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
PTSBench: A Comprehensive Post-Training Sparsity Benchmark Towards Algorithms and Models
PTSBench benchmarks post-training sparsity techniques and model families, finding learning-based allocation and block-wise reconstruction most effective, and attention-based models most sparsity-friendly.