arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2303.09752cs.CLcs.LG
收藏

CoLT5:基于条件计算的更快长程Transformer

CoLT5: Faster Long-Range Transformers with Conditional Computation

Google Research(谷歌研究院)

Joshua Ainslie, Tao Lei, Michiel de Jong, Santiago Ontañón, Siddhartha Brahma, Yury Zemlyanskiy, David Uthus, Mandy Guo, James Lee-Thorp, Yi Tay, Yun-Hsuan Sung, Sumit Sanghai

AI总结:

提出CoLT5模型,通过条件计算为重要token分配更多前馈和注意力资源,在SCROLLS基准上以更快训练推理速度达到SOTA,并有效支持高达64k的极长输入。

摘要:

许多自然语言处理任务受益于长输入,但使用Transformer处理长文档成本高昂——不仅由于平方级注意力复杂度,还因为要将前馈和投影层应用于每个token。然而,并非所有token都同等重要,对长文档尤其如此。我们提出CoLT5,一个基于此直觉构建的长输入Transformer模型,它采用条件计算,在feedforward和attention层中为重要token投入更多资源。我们证明CoLT5以快得多的训练和推理速度取得了比LongT5更强的性能,在长输入SCROLLS基准上达到SOTA。此外,CoLT5能高效且可行地利用极长输入,在高达64k的输入长度上展现出显著收益。

英文摘要:

Many natural language processing tasks benefit from long inputs, but processing long documents with Transformers is expensive -- not only due to quadratic attention complexity but also from applying feedforward and projection layers to every token. However, not all tokens are equally important, especially for longer documents. We propose CoLT5, a long-input Transformer model that builds on this intuition by employing conditional computation, devoting more resources to important tokens in both feedforward and attention layers. We show that CoLT5 achieves stronger performance than LongT5 with much faster training and inference, achieving SOTA on the long-input SCROLLS benchmark. Moreover, CoLT5 can effectively and tractably make use of extremely long inputs, showing strong gains up to 64k input length.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑