paper-with-me

홈 › Papers

Non-Autoregressive Translation by Learning Target Categorical Codes

2021-03-21 · NAACL 2021 4 · Yu Bao, ShuJian Huang, Tong Xiao, Dongqi Wang, Xinyu Dai, Jiajun Chen

Non-autoregressive Transformer is a promising text generation model. However, current non-autoregressive models still fall behind their autoregressive counterparts in translation quality. We attribute this accuracy gap to the lack of dependency modeling among decoder inputs. In this paper, we propose CNAT, which learns implicitly categorical codes as latent variables into the non-autoregressive decoding. The interaction among these categorical codes remedies the missing dependencies and improves the model capacity. Experiment results show that our model achieves comparable or better performance in machine translation tasks, compared with several strong baselines.

📄 PDF Abstract BibTeX arXiv:2103.11405

Code (1)

baoy-nlp/CNAT 공식 구현 pytorch

Tasks

AttributeDecoderMachine TranslationText GenerationTranslation

Methods 이 논문이 사용한 방법론

Linear Layer A Linear Layer is a projection $\mathbf{XW + b}$.
Absolute Position Encodings Absolute Position Encodings are a type of position embeddings for [Transformer-based models] where positional encodings are…
Position-Wise Feed-Forward Layer 설명 없음
BPE Byte Pair Encoding, or BPE, is a subword segmentation algorithm that encodes rare and unknown words as sequences of subword units. The intuition is that various word…
Residual Connection 설명 없음
Layer Normalization Unlike batch normalization, Layer Normalization directly estimates the normalization statistics from the summed inputs…
Adam 설명 없음
Dense Connections Dense Connections, or Fully Connected Connections, are a type of layer in a deep neural network that use a linear operation where every input is connected to every output…

Similar Papers 제목 키워드 기반

Syntactically Supervised Transformers for Faster Neural Machine Translation

2019-06-06 · ACL 2019 7 · Nader Akoury, Kalpesh Krishna, Mohit Iyyer

Standard decoders for neural machine translation autoregressively generate a single target token per time step, which slows inference especially for long outputs. While architectural advances such as the Transformer full…

DecoderMachine TranslationTranslation

MR-P: A Parallel Decoding Algorithm for Iterative Refinement Non-Autoregressive Translation

2022-05-01 · Findings (ACL) 2022 5 · Hao Cheng, Zhihua Zhang

Non-autoregressive translation (NAT) predicts all the target tokens in parallel and significantly speeds up the inference process. The Conditional Masked Language Model (CMLM) is a strong baseline of NAT. It decodes with…

Language ModelingLanguage ModellingTranslation

Integrating Categorical Semantics into Unsupervised Domain Translation

2020-10-03 · ICLR 2021 1 · Samuel Lavoie, Faruk Ahmed, Aaron Courville

While unsupervised domain translation (UDT) has seen a lot of success recently, we argue that mediating its translation via categorical semantic features could broaden its applicability. In particular, we demonstrate tha…

ObjectTranslation

Improving Autoregressive NMT with Non-Autoregressive Model

2020-07-01 · WS 2020 7 · Long Zhou, Jiajun Zhang, Cheng-qing Zong

Autoregressive neural machine translation (NMT) models are often used to teach non-autoregressive models via knowledge distillation. However, there are few studies on improving the quality of autoregressive translation (…

Decoderde-enKnowledge DistillationMachine Translation+3

Retrieving Sequential Information for Non-Autoregressive Neural Machine Translation

2019-06-22 · ACL 2019 7 · Chenze Shao, Yang Feng, Jinchao Zhang, Fandong Meng 외

Non-Autoregressive Transformer (NAT) aims to accelerate the Transformer model through discarding the autoregressive mechanism and generating target words independently, which fails to exploit the target sequential inform…

DecoderMachine TranslationSentenceTranslation