We propose a deep encoder with multiple shallow decoders (DEMSD) where each shallow decoder is responsible for a disjoint subset of target languages.
Jun 5, 2022 · We propose a deep encoder with multiple shallow decoders (DEMSD) where each shallow decoder is responsible for a disjoint subset of target languages.
It has recently been shown for bilingual translation that using a deep encoder and shallow decoder (DESD) can reduce inference latency while maintaining ...
The authors propose a multi-decoder model sharing the same encoder among languages and routing languages in different families to different decoders. ... ...
People also ask
What is multilingual neural machine translation?
What is multi way multilingual neural machine translation?
What is deep neural machine translation?
What is multilingual models for machine translation?
Jun 5, 2022 · Because this deep encoder and shallow decoder model achieves a superior speed-accuracy trade-off on bilingual translation tasks, in this section ...
This work considers several ways to make multilingual NMT faster at inference without degrading its quality, and demonstrates that combining a shallow decoder ...
This paper demonstrates that multilingual denoising pre-training produces significant performance gains across a wide variety of machine translation (MT) tasks.
Multilingual Neural Machine Translation with Deep Encoder and Multiple Shallow Decoders. Xiang Kong, Adithya Renduchintala, James Cross, Yuqing Tang, Jiatao ...
May 29, 2024 · Multilingual Neural Machine Translation with Deep Encoder and Multiple Shallow Decoders. Xiang Kong, Adithya Renduchintala, James Cross ...
Sep 26, 2024 · A deep encoder, shallow decoder architecture, where the encoder retains many layers, but the decoder has fewer, or even only a single layer.