2020/01/13 by Nikita Kitaev, Łukasz Kaiser, Kitaev, Nikita +3 · 176 citations
Computer Science · #Advanced Image and Video Retrieval Techniques #Multimodal Machine Learning Applications #Advanced Neural Network Applications
paper · pdf · doi:10.48550/arxiv.2001.04451
Large Transformer models routinely achieve state-of-the-art results on a number of tasks but training these models can be prohibitively costly, especially on long sequences. We introduce two techniques to improve the efficiency of Transformers. For one, we replace dot-product attention by one that uses locality-sensitive hashing, changing its complexity from O(L2) to O(Llog L), where L is the length of the sequence. Furthermore, we use reversible residual layers instead of the standard residuals, which allows storing activations only once in the training process instead of N times, where N is the number of layers. The resulting model, the Reformer, performs on par with Transformer models while being much more memory-efficient and much faster on long sequences.