paper-with-me

홈 › Papers

Faster Transformer Decoding: N-gram Masked Self-Attention

2020-01-14 · Ciprian Chelba, Mia Chen, Ankur Bapna, Noam Shazeer

Motivated by the fact that most of the information relevant to the prediction of target tokens is drawn from the source sentence $S=s_1, \ldots, s_S$, we propose truncating the target-side window used for computing self-attention by making an $N$-gram assumption. Experiments on WMT EnDe and EnFr data sets show that the $N$-gram masked self-attention model loses very little in BLEU score for $N$ values in the range $4, \ldots, 8$, depending on the task.

📄 PDF Abstract BibTeX arXiv:2001.04589

Code (0)

등록된 구현이 없습니다.

Tasks

Sentence

Similar Papers 제목 키워드 기반

Infusing Sequential Information into Conditional Masked Translation Model with Self-Review Mechanism

2020-10-19 · COLING 2020 8 · Pan Xie, Zhi Cui, Xiuyin Chen, Xiaohui Hu 외

Non-autoregressive models generate target words in a parallel way, which achieve a faster decoding speed but at the sacrifice of translation accuracy. To remedy a flawed translation by non-autoregressive models, a promis…

DecoderKnowledge DistillationTranslation

Mask-Predict: Parallel Decoding of Conditional Masked Language Models

2019-04-19 · IJCNLP 2019 11 · Marjan Ghazvininejad, Omer Levy, Yinhan Liu, Luke Zettlemoyer

Most machine translation systems generate text autoregressively from left to right. We, instead, use a masked language modeling objective to train a model to predict any subset of the target words, conditioned on both th…

Language ModelingLanguage ModellingMachine TranslationMasked Language Modeling+1

M2T: Masking Transformers Twice for Faster Decoding

2023-04-14 · ICCV 2023 1 · Fabian Mentzer, Eirikur Agustsson, Michael Tschannen

We show how bidirectional transformers trained for masked token prediction can be applied to neural image compression to achieve state-of-the-art results. Such models were previously used for image generation by progress…

Image CompressionImage Generation

Accelerating Neural Transformer via an Average Attention Network

2018-05-02 · ACL 2018 7 · Biao Zhang, Deyi Xiong, Jinsong Su

With parallelizable attention networks, the neural Transformer is very fast to train. However, due to the auto-regressive architecture and self-attention in the decoder, the decoding procedure becomes slow. To alleviate …

DecoderMachine TranslationTranslation

Green Hierarchical Vision Transformer for Masked Image Modeling

2022-05-26 · Lang Huang, Shan You, Mingkai Zheng, Fei Wang 외

We present an efficient approach for Masked Image Modeling (MIM) with hierarchical Vision Transformers (ViTs), allowing the hierarchical ViTs to discard masked patches and operate only on the visible ones. Our approach c…

GPUObject Detection