[72] Sparse DETR: Efficient End-to-End Object Detection with Learnable Sparsity

2022λ…„ 10μ›” 20일 Β· 1 λΆ„ Β· long8v Β· 

[71] Large Models are Parsimonious Learners: Activation Sparsity in Trained Transformers

2022λ…„ 10μ›” 17일 Β· 1 λΆ„ Β· long8v Β·