SpAtten: Efficient Sparse Attention Architecture with Cascade Token and Head Pruning
HPCA（2021）
关键词
Natural Language Processing,Attention,Domain-Specific Accelerator,Algorithm-Architecture Co-design,Pruning,Quantization
AI 理解论文
溯源树
样例
生成溯源树，研究论文发展脉络
Chat Paper
正在生成论文摘要