Scalable Transformers for Neural Machine Translation
This addresses the problem of redundant computation and memory in training multiple Transformers for different scenarios in NMT, though it is incremental.
The paper tackles the challenge of deploying Transformers for Neural Machine Translation by proposing Scalable Transformers, which contain sub-Transformers of different scales with shared parameters, and achieves competitive results on WMT EN-De and En-Fr datasets.
Transformer has been widely adopted in Neural Machine Translation (NMT) because of its large capacity and parallel training of sequence generation. However, the deployment of Transformer is challenging because different scenarios require models of different complexities and scales. Naively training multiple Transformers is redundant in terms of both computation and memory. In this paper, we propose a novel Scalable Transformers, which naturally contains sub-Transformers of different scales and have shared parameters. Each sub-Transformer can be easily obtained by cropping the parameters of the largest Transformer. A three-stage training scheme is proposed to tackle the difficulty of training the Scalable Transformers, which introduces additional supervisions from word-level and sequence-level self-distillation. Extensive experiments were conducted on WMT EN-De and En-Fr to validate our proposed Scalable Transformers.