arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2103.14636cs.LG
收藏

更快更轻的Transformers的实用化研究

A Practical Survey on Faster and Lighter Transformers

Polytechnique Montréal(蒙特利尔理工学院);Mila - Quebec AI Institute(米拉-魁北克人工智能研究所)

Quentin Fournier, Gaétan Marceau Caron, Daniel Aloise

摘要:

递归神经网络是处理序列的有效模型。然而,由于其固有的顺序性,它们无法学习长期依赖性。作为解决方案,Vaswani等人引入了Transformer,这是一种完全基于注意力机制的模型,能够将输入序列的任何两个位置联系起来,从而对任意长的依赖关系进行建模。Transformer改进了众多序列建模任务的最新技术水平。然而,其有效性是以相对于序列长度的二次计算和存储复杂度为代价的,这阻碍了其采用。幸运的是,深度学习社区一直对提高模型的效率感兴趣,导致了大量的解决方案,如参数共享,修剪,混合精度和知识蒸馏。最近,研究人员通过设计较低复杂度的替代方案,如Longformer、Reformer、Linformer和Performer,直接解决了Transformer的局限性。然而,由于广泛的解决方案,它已成为具有挑战性的研究人员和从业人员,以确定在实践中应用的方法,以满足所需的容量,计算和内存之间的权衡。本调查通过调查使Transformers更快、更轻的流行方法来解决这个问题,并对这些方法的优势、局限性和基本假设进行全面解释。

英文摘要:

Recurrent neural networks are effective models to process sequences. However, they are unable to learn long-term dependencies because of their inherent sequential nature. As a solution, Vaswani et al. introduced the Transformer, a model solely based on the attention mechanism that is able to relate any two positions of the input sequence, hence modelling arbitrary long dependencies. The Transformer has improved the state-of-the-art across numerous sequence modelling tasks. However, its effectiveness comes at the expense of a quadratic computational and memory complexity with respect to the sequence length, hindering its adoption. Fortunately, the deep learning community has always been interested in improving the models' efficiency, leading to a plethora of solutions such as parameter sharing, pruning, mixed-precision, and knowledge distillation. Recently, researchers have directly addressed the Transformer's limitation by designing lower-complexity alternatives such as the Longformer, Reformer, Linformer, and Performer. However, due to the wide range of solutions, it has become challenging for researchers and practitioners to determine which methods to apply in practice in order to meet the desired trade-off between capacity, computation, and memory. This survey addresses this issue by investigating popular approaches to make Transformers faster and lighter and by providing a comprehensive explanation of the methods' strengths, limitations, and underlying assumptions.

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。

↑