Large language models (LLMs) have revolutionized natural language proces...
This paper presents a Spatial Re-parameterization (SpRe) method for the ...
Token compression aims to speed up large-scale vision transformers (e.g....
Arbitrary bit-width network quantization has received significant attent...
CutMix is a vital augmentation strategy that determines the performance ...
We attempt to reduce the computational costs in vision transformers (ViT...
Vision Transformers (ViT) have made many breakthroughs in computer visio...
Network sparsity receives popularity mostly due to its capability to red...
While post-training quantization receives popularity mostly due to its
e...