Wang, Yaqing
18 publications
NeurIPS
2024
FIARSE: Model-Heterogeneous Federated Learning via Importance-Aware Submodel Extraction
ICML
2024
Outlier Weighed Layerwise Sparsity (OWL): A Missing Secret Sauce for Pruning LLMs to High Sparsity
ICLRW
2024
Outlier Weighed Layerwise Sparsity (OWL): A Missing Secret Sauce for Pruning LLMs to High Sparsity
MLJ
2023
Distilling Ensemble of Explanations for Weakly-Supervised Pre-Training of Image Segmentation Models