TETRIS: TilE-matching the TRemendous Irregular Sparsity

Yu Ji, Ling Liang, Lei Deng, Youyang Zhang, Youhui Zhang, Yuan Xie

Neural Information Processing Systems 

Compressing neural networks by pruning weights with small magnitudes can significantly reduce the computation and storage cost. Although pruning makes the model smaller,itisdifficult toget apractical speedup inmodern computing platforms such as CPU and GPU due to the irregularity.

Similar Docs  Excel Report  more

TitleSimilaritySource
None found